We benchmarked the RTX PRO 6000 in gaming scenarios using an AMD Ryzen 9800X3D, experimented with LLM benchmarks, and ran thermal and acoustic tests

The Highlights

• The RTX 6000 features 24,064 CUDA cores, which is nearly an 11% increase over the RTX 5090
• Unlike the 5090, which uses liquid metal, the RTX Pro 6000 uses thermal paste
• On the gaming side, the RTX Pro 6000 outperformed the 5090 by roughly 5 to 14% in our tests, but that alone isn’t a good reason to buy the card
• Original MSRP: $8,000 - $11,000
• Release Date: March 18, 2025

Table of Contents

• AutoTOC

Intro

We paid $8,000 for an RTX PRO 6000 GPU with 96GB of VRAM when we bought it from Brent Rambo, the actual guy in this meme, who’s grown up now and had access to an RTX Pro 6000. These days, Rambo is busy with custom and bespoke PC builds these days on his own website, and that means he has access to interesting hardware.

It’s a shame that NVIDIA has left us feeling like it’s holding its engineers hostage for manipulating reviews, but it’s clear that we don’t mind buying our own hardware to review.

Editor's note: This was originally published on June 24, 2025 as a video. This content has been adapted to written format for this article and is unchanged from the original publication.

Credits

Test Lead, Host, Writing, Editing

Steve Burke

Testing

Patrick Lathan

Video Editing

Vitalii Makhnovets

Camera, Video Editing

Tim Phetdara

Testing, Writing, Camera

Tannen Williams

Writing, Web Editing

Jimmy Thang

The RTX PRO 6000 is the best marketing for AMD’s Ryzen 9800X3D (read our review) that we use in our GPU test benches, mostly because we’re seeing 5% to 14% scaling in gaming benchmarks versus the RTX 5090 (read our review). We also tore it down and found the most insanely dense PCB we’ve ever seen with a reversion away from liquid metal.

But this doesn’t have 96GB of VRAM just because that’s where it all went when they lost it on the RTX 5080 (read our review), 5070 Ti (read our review), 5070 (read our review), 5060 Ti 8GB (read our review), and 5060 (read our review)-- it’s also there because it’s better for... AI.

Today, we’re benchmarking some of those in a round of experimental benchmarks. We haven’t done machine learning and LLM tests before and we don’t have a full methodology defined yet, but we’ve begun experimenting.

NVIDIA RTX 6000 Overview

This card is not for gaming. That much is obvious. We’re still testing it in games, and that’s mostly for our own purposes: We want to know how much room the 9800X3D has in it to scale for GPUs, and fortunately, it looks like there’s still plenty for another generation of reviews. The 9800X3D keeps up with the RTX PRO 6000 even when at 1080p in some situations. 

The biggest difference between the PRO 6000 and the 5090 after the 96GB vs. 32GB memory capacity is the GPU itself: The PRO 6000 has 24,064 CUDA cores to the 21,760 of the 5090, nearly an 11% increase. That’s a big difference.

There are multiple versions of the RTX 6000. The one we're looking at specifically is called the Blackwell Workstation Edition, but NVIDIA also has so-called “server editions” of the cards, and those are intended to go into servers and racks where they're getting force fed air from the front of the chassis rather than using an FE style cooler, which is the one we're looking at today. So, the one we're looking at is more similar to what we would see in sort of the consumer to prosumer workstation class.

NVIDIA RTX 6000 Specs, Architecture Basics, and Price

Getting into the specs:

The RTX PRO 6000 Blackwell GPU is built on the GB202 Blackwell die. The full GB202 die has support for up to 24,576 CUDA cores, 768 TMUs, 192 ROPs that might even be present, and is also used for the RTX 5090.

The PRO 6000 variation on this GPU has 24,064 CUDA cores, so it’s down by 512 and isn’t a perfect die, with 752 TMUs (down from 768) and still 192 ROPs, and they’d better all be there for the price. The card is also running 188 RT cores and supports PCIe 5.0 x16. Advertised clocks are 1590 MHz base and 2617 MHz boost. The real reason people buy this card is its memory, though, at 96GB of GDDR7.

The closest consumer class card to compare against is the RTX 5090. It’s built on the same die, but with 176 ROPs, 680 TMUs, 170 RT cores, 21,760 CUDA cores, 32GB of GDDR7 memory, a 512-bit bus width, and a far lower price at $2,000 to, realistically, $3,000 rather than $8,000 to $11,000.

For some notable differences: The 5090 only has a third of the memory capacity of the PRO card, and the RTX PRO 6000’s die is almost 98% present while the RTX 5090’s die is closer to 88.5% present, with the remainder either being fused off or defective fallout and fused off. 

You could also technically buy four 5090s for the price of one RTX PRO 6000, so the use case really does come down to having as much memory as possible attached to a single GPU and PCIe slot.

NVIDIA RTX 6000 Tear-Down

The RTX Pro 6000 we tested is a 2-slot card with 2 large fans on the front. The PCB is in the center of the card and you can see where the fins start to get taller. This indicates where the PCB plate ends. 

The card’s PCIe slot is connected via a customized pinout adapter. 

If you look at the backside of the card where the fans are, you’ll notice a slight depression down towards the middle of the fins. This is supposed to help with pressure drop. The design is full flow-through, which pushes air straight through the card rather than into a PCB. 

The other side of the card looks very similar to a 5090. 

The top of the RTX Pro 6000 retains the angle on the card’s vented slats. This projects the air out at an angle that’s up and away from the card. 

Like the 5090, the card uses a 12VHPWR cable.

Starting disassembly, we removed the card’s back plate. Then we removed a single screw, which allowed us to pull off the card’s exterior frame.  

From here, we removed some screws around the center of the card to remove the center covers. Doing this reminded us of how masterful the mechanics of the FE design is. It’s assembled really well, it’s easy to take apart, and there’s no mechanisms to try and prevent disassembly.   

Next, we unscrewed the clamp that holds the PCIe slot to the card.

Here’s a close look at the removed PCIe adapter. 

From here, we pried a plate off of the card, exposing the PCB. 

This gave us our first look at where the memory is connected. We counted 16 memory modules on the back with the clay type of thermal pads. 

The PCB is crazy dense, which is an engineering challenge. 

We then proceeded to remove ribbon cables and screws to the leaf spring to access the GPU. 

Pulling off the PCB, we were expecting to see liquid metal, but we saw thermal paste instead. We really weren’t expecting that. We also saw a massive GPU substrate.

Upon removing the PCB, we noticed that 1 pad went over the edge a bit. 

Analyzing the paste application pattern, it looked like the heaviest imprint was right in the center with lighter pressure on the outer edges, which is fine.  

Looking at the thermal pads, they were making clear contact, though we did see some pads that were smashed over the sides a bit. This is fine, however. We counted 32 total memory locations, which means that they are 3GB memory modules as that totals 96GB, the capacity of the card. 

Taking a closer look, we noticed some areas of poor contact along one edge of the inductors, where we noticed that half of the inductors are actually covered by a thermal pad. One of them only had about a third of the inductor covered. That’s not good as they really need to get the pad placement correct on GPUs that cost 8-11 thousand dollars.

The board also has lots of little tiny components, leading to a very dense PCB.  

Finally, we cleaned off the thermal paste off of the GPU, exposing the GB 202-870-A1 SKU. This is a 5090 die, just a fuller version of it.  

NVIDIA RTX 6000 Thermal Benchmarks

Thermal testing is up now. The RTX 5090 uses liquid metal with a 2-slot FE cooler that, genuinely, is one of the most impressive GPU coolers we’ve ever worked on. The prototype 4090 cooler we disassembled was more impressive in mechanical and cooling capabilities, but it was huge and impractical to manufacture. The 2-slot FE has done well in most cases, with the exception being memory cooling. We found that memory ran hot on our 5090 FE. That’s more of a concern with the PRO 6000 as well since it has so much memory.

Here’s a thermal test during a fixed render workload. The 5090 ran at around 72 degrees Celsius for the GPU core temperature in this test, which is overall excellent considering its 2-slot design. 

The VRAM for the 5090 ran at around 90 degrees Celsius, which was hotter than we’re comfortable with. Once accounting for installation into a case and potential long-term implications of pad dryout and dust, that’s high. The PRO 6000 ran its GPU core at a significantly hotter 82 degrees Celsius in this test, so about 10 degrees warmer than the 5090. The GPU memory was about the same, measuring 88 degrees via software.

Fan Response

None of that means anything without fan speeds.

Using only the auto settings, so following whatever VBIOS has programmed for the target GPU temperature, the RTX 5090 ran at 1550-1600 RPM for the average fan speed. The PRO 6000 ran at 1700. That’s a little faster with a much warmer core, but similar memory thermals.

Acoustics

Acoustics are next. For this testing, we’re using our hemi-anechoic chamber that we heavily invested in for improving our test quality. Our next major improvement will be upgrading the microphone, which is currently our bottleneck. We could hit a 6-8 dBA noise floor with a better mic, but we’re currently at 13.6 to 15.0 dBA for the floor. That’s still great though.

The RTX 5090 ran at about 32.5 dBA, with the PRO 6000 at the same noise level; however, the noise floor was slightly different between these two, so adjusting for that, the PRO 6000 would be about 1 dBA louder under the same conditions. This is hardly noticeable as a difference, if noticeable at all for most people.

Overall, the cooler design remains good when taking its size into consideration. The 5090 had higher noise levels in the higher frequency range, particularly 8000 to 10000 Hz. It also had a peak around 175 Hz, with the PRO peaking at 194 Hz and following a similar pattern, just adjusted right.

Generally speaking, this mostly follows the same trend. Our frequency cutoff is 150 Hz, so the higher result on the left can be ignored.

NVIDIA RTX 6000 Gaming Benchmarks

Gaming benchmarks are next. 

Dragon’s Dogma 2 - 4K

Dragon’s Dogma 2 is first. 

Tested at 4K, the RTX PRO 6000 landed at 140 FPS AVG, which has it ahead of the RTX 5090 by 5.8% for average framerate. Lows are not notably different. In other words, that’s about an extra $1,000 per 1% improvement, or about $780 per 1 FPS increase over the 5090 when calculating by the difference in cost for these two.

The lead in the RTX PRO 6000 is about 42% over the 4090 (watch our review), with the 5090 already ahead of the 4090 by about 34% here.

Dragon’s Dogma 2 - 1440p

At 1440p, the lead over the 5090 is expanded to 6.3%, ensuring we get our full gaming value out of the card. Low performance is not different outside of the usual scaling along with the improvement in average frametimes.

The RTX PRO 6000 hits 201 FPS AVG here, leading the 189 FPS of the 5090 and the 155 of the 4090. NVIDIA currently holds the entire top cluster of this chart, with AMD mostly focusing on the modern mid-range and Intel focused on the modern low-end.

Dragon’s Dogma 2 - 1080p

We next tested 1080p, which is what we all know buyers of this card really want to use it for. So-called AI use cases are obviously secondary compared to playing video games at 1080p.

The RTX PRO 6000 leads this extremely important and pivotal chart with a 226 FPS AVG, dropping to just a 5.3% lead. Honestly, there is something valuable that comes out of this chart: It’s marketing for AMD’s Ryzen 7 9800X3D, which is somehow keeping up with these GPUs enough that it can still produce distinguished results between the $2,000-$3,000 5090 and the $8,000 to $11,000 PRO 6000. That’s impressive. This is more amusing to us as it proves the longevity of our bench hardware more than anything else.

FFXIV 4K

Final Fantasy 14 is up now, tested first at 4K.

The RTX Pro 6000 ran at 208 FPS AVG, leading the 5090’s 184 FPS AVG result by about 13%. That’s a larger gain than we’ve seen in some of the other games, though obviously this is still a card intended for VRAM-intensive use cases and not gaming. That there’s still some power left beyond the 5090 is what’s more interesting, alongside the fact that the 9800X3D is so capable. Lows scale with the average and are not meaningfully better with the PRO 6000.

FFXIV 1440p

Tested at 1440p, we see a 10% improvement in framerate against the 5090, up at 348 FPS AVG for the 6000. Again, we’re mostly excited to see the 9800X3D continuing to scale and illustrating that there’s room left for another generation of testing on these benches. Lows remain in-step with the average.

FFXIV 1080p

At 1080p, the PRO 6000 again outperforms the 5090. This time, it’s reduced to about 4.3%. Let’s move on.

Starfield - 4K

Up next, we’re testing a card no one buys for gaming with a game no one plays. In Starfield at 4K, the PRO 6000 ran at 115 FPS AVG, leading the 5090 by about 7%. Lows are mostly within error. The value here is somewhere in the range of $700 to $1,000 per 1% improvement, depending on the price of the 5090.

Starfield - 1440p

At 1440p, the RTX PRO 6000 ran at 154 FPS AVG, leading the 147-148 FPS result of the 5090 marginally. The improvement over the 4090 is 17%, with the 5090 having previously led the 4090 by 12% in this test, as we’re becoming bound elsewhere.

The 1% lows for the 6000 are noteworthy here: At 89 FPS for the averaged 1% low, the PRO 6000 outperforms the average framerate of the 5070 and nearly matches the 4070 Ti with just the PRO’s 1% numbers alone. 

Resident Evil 4 - 4K

Resident Evil at 4K is up next. In this one, the RTX PRO 6000 ran at 219 FPS AVG, leading the 207 FPS AVG result of the 5090 by 6%. That’s consistent with most of the other tests so far. The lows, again, are not meaningfully different. The 4090’s 151 FPS AVG allows the 6000 a lead of 45%, or the 5090 a lead of 36.7%.

Resident Evil 4 - 1440p

At 1440p, the Blackwell workstation card ran at 371 FPS AVG, pushing the framerate measurably higher than the roughly 350 FPS AVG of the 5090 previously. We’re looking at about a 6.5% improvement, aligning with prior results again, and yet again showing just how good the 9800X3D is. 

Black Myth: Wukong - 4K

Black Myth: Wukong is up now, tested first at 4K and representing one of our heaviest non-RT workloads. The RTX PRO 6000 didn’t break the 100 FPS barrier, but it was the first card to exceed 90 -- not that this particularly matters to anyone when the 5090 was already in the mid 80s. The 6000 outperforms our 5090 FE result by 7.4%, above what we’ve seen on average thus far by about 1 percentage point. 92 FPS AVG is a good showing in this benchmark when at 4K. The next card in the stack is the 4090 at 67 FPS AVG, then the 5080 at 58 FPS AVG.

Black Myth: Wukong - 1440p

At 1440p, the PRO 6000 ran at 138 FPS AVG and led the 5090 marginally, but it was technically still a measurable improvement. Total uplift is 6.1%, aligning with Resident Evil 4 previously. Once again, the next card is the 4090. After this is the 5080, which isn’t that different from the 4080 Super (read our review), which isn’t that different from the 4080 (watch our review), which isn’t that different from the 5070 Ti. 

Black Myth: Wukong - 1080p

Black Myth: Wukong at 1080p is next. Selfishly, 1080p is interesting once again for probing at if there’s room in the CPU for more GPU. Turns out, there is: With the 9800X3D, the RTX PRO 6000 ran at 168 FPS AVG, improving on the 5090 by about 5%. That’s less than we saw at 1440p, which was less than we saw at 4K, so the card is gaining ground as resolution increases.

Dying Light 2 - 4K

Dying Light 2 is up next. At 4K, the PRO 6000 ran at 144 FPS AVG, which is a noteworthy improvement on the 126 FPS AVG of the 5090 FE. That’s a surprising 13.8%, which so far matches only one other game we’ve tested -- and that was Final Fantasy. Most of the other games are closer to 5-7%.

The 4090 is next at 91 FPS AVG, meaning the PRO 6000 is 59% ahead of the 4090.

Dying Light 2 - 1440p

At 1440p, there’s still scaling: The PRO 6000 is now at 246 FPS AVG, leading the 5090 by almost 13%. That’s a drop from what we saw at 4K, but still more than most other games.

Cyberpunk 2077: Phantom Liberty - 4K

Cyberpunk 2077: Phantom Liberty is one of our next most intensive games after Black Myth: Wukong. At 4K, the RTX PRO 6000 breaks through 100 FPS and hits 108, leading the 5090 by 13.5%. This is actually a pretty big lead and is another one of the games that has a larger average improvement. That’s still not worth $8,000 or $11,000 or whatever it is they’re selling these for, but it’s interesting to see any difference at all given the focus on so-called AI workloads.

Cyberpunk 2077: Phantom Liberty - 1440p

At 1440p, the PRO 6000 ran at nearly 200 FPS AVG, about 9.4% ahead of the 5090. This remains one of the larger gaps for 1440p, although again, it’s obviously not something anyone should remotely consider primarily for gaming. The 4090 is down at 135 FPS AVG, meaning the PRO 6000 leads it by 48%. 

Cyberpunk 2077: Phantom Liberty - 1080p

We’re leaving 1080p in because it’s the only place we see a bottleneck. Finally, we run into what seems like a CPU limit. The PRO 6000 and 5090 are at about the same level of performance, as is the 4090. Everything here is limited. That means our 1440p results were at least somewhat limited as well, with minimally the upper bound bouncing off of limits.

NVIDIA RTX 6000 Ray Tracing Benchmarks

Ray tracing benchmarks are next.

Ray Tracing - Black Myth: Wukong 4K

We’ll start with the heaviest one, which is Black Myth: Wukong at 4K with upscaling.

In this test, the RTX PRO 6000 ran at 92 FPS AVG, which was barely any different from the 88 FPS observed on the 5090. We’re under 5% of uplift here. It’ll be interesting to see if RT workloads don’t produce as much change as rasterization.

Ray Tracing - Black Myth: Wukong 1440p

At 1440p upscaled, the RTX PRO 6000 ran at 131 FPS AVG, leading the 5090 by just 3%. There’s just not much change in this test in general.

We’ll skip 1080p given the lack of improvement.

Ray Tracing - Cyberpunk 2077 (4K, RT Medium)

Cyberpunk at 4K with RT Medium is next. In this test, the RTX PRO 6000 ran at 65 FPS AVG, leading the 5090 by 11%. That’s roughly in-line with what we saw in Cyberpunk rasterized. This is without any upscaling, which may be helping the PRO card’s relative gain.

And finally, unlike what we saw in the 5060 and 5070 class cards with significant VRAM limitations harming low performance, the PRO doesn’t have any VRAM issues. This is clearly the solution. The solution is to buy $8K-$11K video cards.

As before, note that some cards with low VRAM, like the 5070 and 4070 Ti, have averages that look better than the reality (even though the averages are also not great).

Ray Tracing - Cyberpunk 2077 (4K, RT Ultra)

4K with RT Ultra is next. We almost never publish this test since it’s so intensive that it becomes somewhat useless, but it’s interesting here.

The RTX PRO 6000 ran at 56.6 FPS AVG, leading the 5090 by 6.4%. That’s a significant drop from what we saw at RT Medium, indicating that the increase in intensity for the RT workload is minimizing the benefit of other aspects of the PRO 6000’s improvements. It is getting overrun by the RT workload intensity. The lead over the 4090 is about 44% here for the 6000, with the 5090 leading the 4090 by 35%.

Note that the lower portion of this chart is unreliable for average framerate since the cards are all unplayable and stuttering. This can sometimes be due to exceeding VRAM limitations, such as on the RTX 5070. The average looks far better than the reality, even though the average is also unplayable.

Ray Tracing - Cyberpunk 2077 (1080p, RT Medium)

At 1080p with RT Medium, the PRO 6000 ran at 182 FPS AVG with lows paced proportionally. That has it ahead of the 5090 by about 12 FPS AVG, or about 7%.

Ray Tracing - Dragon’s Dogma 2 4K

Dragon’s Dogma 2 with ray tracing and without upscaling is next. Tested at 4K native, the RTX PRO 6000 at 121 FPS AVG leads the 5090 FE by about 7% in average framerate. This is starting to match a pattern, so we’ll let this be our last game test. Let’s move on to something else.

RTX 6000 AI Benchmark Charts:

Now, we'll get into some of our first ever LLM and machine learning benchmarks. As a disclaimer, we know enough to get some charts together to run some benchmarks where we feel like the controls of the test environment are good, but we do not yet know enough to have a full picture of capabilities outside of the test suite that we're using currently. 

We've tested with LM Studio, 3DMark testing, and ML Perf. We kind of settled on one set just to experiment for now. And this is our disclosure that we are experimenting with new testing, which means that the charts get the experimental chart label on top of them. And that's just so everyone's aware that these are not up to our full confidence standards of our normal benchmarks that we do all the time because we're still learning here.

LM Studio Testing

For our LM Studio testing, we recorded the response speeds from 8 different models that incrementally increase in size. 

We’ll get to larger models in a second.

Our first chart illustrates the AVG tokens/second results from the three smaller models – ranging from 8.5GB to around 19GB in size. We think these first three are more representative of like-for-like tests compared to some other models we’ll be going over today – because any of the GPUs on our chart can load these models without being limited by VRAM capacity.

In DeepSeek Llama 8B Distil, the RTX PRO 6000 is functionally tied for first place with the RTX 5090 FE – both with response speeds of roughly 81 tokens/second, improving on the 4090’s 62 tokens/second AVG by about 30%.

In 8-bit Phi-4, the PRO 6000 puts some distance between its own tokens/second AVG of 62 and the 5090's 51 tokens/second AVG, showing a 22% improvement. Compared to the 4090, the RTX PRO improves by about 43%.

In Qwen 2.5, the PRO 6000's 44 tokens/second AVG sees improvements of 25% over the 5090 and 37% over the 4090.

Text Generation

Moving to our next chart, we’ll really start to see performance gaps begin to widen as VRAM limitations take effect.

The RTX 6000 tops the chart in all five models.

In InternLM, the RTX PRO 6000's 50 tokens/second AVG improves upon the 5090's 40 by 25% and upon the standard 24GB 4090 by 319%.

In Mistral Small 26GB, the 4090 and 5090 see much sharper performance decreases than the PRO 6000. The workstation GPU achieves 42.4 tokens/second AVG, or 147% improved from the 5090’s 17 and 560% greater than the 4090’s AVG of 6.4.

Once we reach Gemma 3 27B, the workstation card really starts to separate itself from the lowly $2,000 gaming cards. In this model, the RTX PRO 6000 achieves a 29 tokens/second AVG, with the 5090 only reaching 5 and the 4090 only seeing 4 tokens/second, give or take some change. The comparisons remain largely the same in QwQ 32B.

In the final Llama 3.3 70b Q4_K_S model we tested, the RTX PRO 6000 achieved its greatest lead yet. The workstation card sees improvements of 928% over the 5090 and 1141% over the standard 4090. 

A more simplistic way to interpret this data is: At a certain point, the other cards run out of VRAM, while the workstation GPU doesn’t even use half of its VRAM in the largest Llama model we tested. You could expand beyond this as well, of course, but that’s the extent of our experimentation for now.

NVIDIA RTX 6000 Conclusion

That's it for the RTX Pro 6000 test. It's a review in a way but at the same time, we'd want to do a lot more tests in the ML category to really have a fully-fledged review. This piece has been an exciting experiment for us. 

The biggest discovery for us was seeing no liquid metal while we were doing the tear-down. That was probably known, but it’s not something that we had looked into. So that was interesting. We could see why NVIDIA did that for reliability reasons. We could also see it for liability reasons if the company is worried about liquid metal leaking out onto expensive servers. We kind of doubt this was done to save on costs, though maybe it was done for liability costs, but it just doesn't seem like NVIDIA is going to try and save a buck on this kind of card. 

We spoke to Wendell from Level1Techs and he noted that there's some apparent buggy behavior with Blackwell right now and so that's still getting updated and perhaps there's performance optimizations that could be done. On the side that we're familiar with, which is the gaming aspect, seeing 5 to 14% improvement is mostly useful not because you should ever in any way consider this card for gaming, because that would be an insane waste of money if that's all you would do with it, but more because it shows us that there's room left in the 9800X3D for scaling, which is pretty cool.

We compare Yeston’s “Waifu” RX 9070 XT vs PowerColor’s Red Devil equivalent in a series of tests that include acoustics, frequency, temps, gaming performance, and more

The Highlights

• Both RX 9070 XT cards cost $900
• The Red Devil 9070 XT needs to better tune its fan curves
• Both cards aren’t bad, but are overpriced at $900

Table of Contents

• AutoTOC

Intro

This is the most important review of all year. This is what makes or breaks this industry -- and even history: On one side, we have anime waifus from Atlantis bathed in the pleasant scent of the ocean vying for neckbeard attention, and on the other side, we have the Hellstone powering the Devil’s edgelord doomsayer RX 9070 XT. Only one will emerge victorious.

Editor's note: This was originally published on July 9, 2025 as a video. This content has been adapted to written format for this article and is unchanged from the original publication.

Credits

Test Lead, Host, Writing

Steve Burke

Testing

Mike Gaglione

Camera

Vitalii Makhnovets
Tim Phetdara

Animation, Editing

Andrew Coleman

Writing, Web Editing

Jimmy Thang

The Yeston RX 9070 XT Sakura Sugar Atlantis OC Edition weighs in at $900 and has its own YouTube channel devoted to Sakura, the angel of the ocean, including power move footage of soaking a video card in the very ocean sands whence it came. 

And when Yeston finally releases its much anticipated Husbando video card, named after the cool and aloof character “GAME ACE,” it’s GAME OVER for PowerColor’s Red Devil.

But Yeston isn’t alone in this battle of gamer marketing: PowerColor’s Hellhound says it has a “New Dawn” and “Power in Every Shade,” And... we don’t know what that means.

But it sounds scary -- almost as scary as the RED DEVIL, the natural counter to the WAIFU. The Red Devil says its Hellstone is “born from darkness, radiating power,” “forged in the fires of the underworld and inspired by the cradle of the rarest gems.” PowerColor “emerges as the unparalleled force in gaming, pushing performance to its absolute peak,” and “at its core lies the Hellstone, a breathtaking design element where vibrant RGB pulses with an otherworldly glow, embodying fiery power and precision.” 

Such a stone sounds too powerful to be in the hands of mere, filthy casuals. The breathtaking design element Hellstone is made of hundred-million-year-old materials dug from the depths of the earth. We’re told that it’s made of a rare, nearly unattainable polymeric matrix composite from hydrocarbon polymers, known only to mortal humans as... “plastic.”

This is the battle you all have been waiting for. It is the ultimate GPU sh*tpost -- second only to the one that NVIDIA just made by launching the so-called RTX 5050 (watch our coverage).

And if you’re wondering if this is just one gigantic meme with actual testing that required dozens of hours of work and hundreds of thousands of dollars worth of equipment all for this one meme, then yes. Yes it is. But it does have an important conclusion. And we are going to get to the bottom of who is better between the RED DEVIL VS. the WAIFU. Let’s battle. 

Testing

Despite the jokes, we do actually have testing in our review. We’re testing thermals, some power, acoustics, frequency, and a couple of games.  

The main reason we’re looking at games is because the 9070 XT Red Devil is running at a pretty high clock compared to other 9070 XTs so we wanted to look at that to see how much it actually mattered. 

The biggest issue with both cards is that they are currently $900, which is completely insane because MSRP is allegedly $600 for the 9070 XT, though that’s not really the case. Regardless, $900 is still pretty high. 

Both the cards we’re reviewing here have dual vBIOS and share a lot of common features. This means we’re going to focus most on aspects that are likely to be different. This includes thermals and acoustics. 

Thermals

The first test of prowess for these embattled icons is one of thermals: The Hellstone has a clear disadvantage here, mostly because it’s, uh, from hell, and the Waifu has an advantage with its affinity for oceans and Atlantis.

For this testing, our methodology involved testing the auto out-of-box settings in addition to testing with noise-normalized manual overrides to ensure an even battlefield. Waifus and Devils make different noises at different volumes, after all, and so it’s only fair to normalize them. 

The normalization required use of our hemi-anechoic chamber that we had custom-built exactly for purchases like this. You might think that spending $250,000 on a chamber to test the audio quality of meme video cards is insane, and you’d be right. That’s why we also test non-memes in it and use it every single benchmarking and technical reviews that we do for coolers and cases. To support our ability to buy test equipment like this and provide accurate data, head over to store.gamersnexus.net and grab one of our unique 3D Coaster Packs, like from our Debug pack. These have carefully placed and tested components to stabilize drinkware while providing a soft, rubberized material with PC theming. The debug code and power and reset switches on the red-and-black motherboard is one of our favorites, although the original coaster pack (also on the store) has the fan coaster that works surprisingly well and is also the most popular. Each pack comes with 4 coasters. Despite price increases over the years, we have maintained our original price. Head over to store.gamersnexus.net to grab these and support our in-depth testing and reviews -- even for memes.

Let’s get into the tests.

Thermals - 3DM 4K Auto VBIOS

The Waifu starts out with a strong showing when set to follow the default VBIOS profile and temperature targets with auto fan controls. The Atlantis-dweller ran at 57 degrees Celsius at steady state in a controlled ambient setting for GPU core on both the main and alternate VBIOS; in fact, as we showed in our Yeston Waifu review, the VBIOSes appear to do the same thing. It’s like they didn’t reprogram one of the fan curves or something, which may be a mistake.

But these temperatures are completely acceptable. They are maintained at 23.6 dBA and about 304W via GPU-Z board power monitoring.

The PowerColor Red Devil runs hotter: Its default OC VBIOS held a 60.7-degree result, with the silent VBIOS also seemingly the same. This has happened for multiple generations now with both Yeston and PowerColor. Maybe because the load hits steady state, they end up settling in the same place once under sustained load. We are restarting between each VBIOS change and even inspect the VBIOS profile name when switching to confirm that it does toggle. This process works to show differences in other devices, but at least here, we’re seeing the same across the profiles. That at least keeps it simple.

The Red Devil is running about 3-4 degrees warmer on the GPU and about 1-2 degrees warmer on the hotspot, with memory at a substantial climb of 8 degrees warmer. There’s a deficiency in memory cooling on the Red Devil by comparison, but an advantage in the hotspot-to-core delta, where Yeston’s delta is greater.

But there’s one other important factor: Out of the box, the PowerColor card is pulling 30W more via GPU-Z logging. This puts them at more equal thermal footing when considering the increased power consumption. The fan speed is another huge factor. So that was out of the box thermal performance, but we can start controlling these variables.

Thermals - 3DM 25 dBA

This test normalizes the GPUs for noise levels. We ran both at 25 dBA at 1-meter for this test.

The steady state 3DMark workload has the Waifu now 6 degrees warmer than the Red Devil in a searing blow to the Waifu, whose ocean and its scent-imbued plastic is evaporating before our very eyes. Boosting the Red Devil’s fan speeds to match the noise levels of the Waifu’s brings it down significantly, showing that it is the better card in like-for-like acoustic testing. Even with its smoldering Hellstone, the Red Devil has a 76-degree hotspot and delta of 26 degrees, improved from the Waifu’s 84-degree hotspot delta of 27.8-degree delta. This means the waifu is hotter here. Memory thermals are functionally tied between them, showing that PowerColor’s main area to improve is VRAM thermals. All of these numbers are acceptable, though, with the GPU numbers just straight-up good for both companies.

Notably, PowerColor is achieving these results with higher power draw. It lost when auto due to poor fan configuration, but pulls ahead when manually normalized.

Thermals - FurMark 25 dBA

In FurMark at 25 dBA, which loads the GPUs more heavily on the VRM, the GPU temperatures ended up comparable between these with the memory temperatures favoring the Waifu by about 2 degrees reduced from the Red Devil.

Acoustics

But this doesn’t mean much without acoustics. 

The most natural and first test of a Devil against a Waifu is the noise that they make in battle. For this reason, we threw these two GPUs into our battle arena: The hemi-anechoic chamber.

The Yeston RX 9070 XT Waifu edition ran at 1,130 RPM when under default VBIOS control at 304W in the workload that we ran. The card ran at 23.6 dBA at 1 meter when tested under the conditions we used for thermal testing. The frequency spectrum analysis has the highest spike around 164 Hz, with a swell around 500 Hz and a later bump and dip in the range of 1,000 Hz to 1,400 Hz.

The PowerColor Red Devil ran at 16.8 dBA, which is really quiet for what we would imagine of hell. It’s too close to our floor to be all that useful. The Devil is substantially quieter out of box, mostly because it just isn’t utilizing its fans as well as we think it should be -- that’s why it ran warmer in the auto testing. We did some in-depth investigating and realized it’s because the Hellstone converts the Red Devil’s angry howlings into the power needed to sustain its own power consumption, destroying the sound waves in the process. Fans ran at 910 RPM rather than 1,130 RPM, which means it’s quiet, but we think also not properly utilized. PowerColor could find more of a balance here, although maybe they’re trying to remain true to the fiery marketing.

The frequency spectrum has the Devil’s peak at around 320 Hz, with an elevated area around 420 Hz. Overall, it’s a lot quieter than the Waifu, with the high-end of the frequency scale similar.

Acoustics - Noise-Normalized

When normalized to roughly the same noise levels, leaving the Yeston card on the chart, the Red Devil’s new frequency spectrum has the Devil’s peak at about 200 Hz. There’s a slight wave up around 400-500 Hz and another around the same place as the Yeston card, at 900 Hz, then a dip, then up again at around 1,200 Hz.

Frequency

GPU frequency is the next test in this battle of the memes. The Yeston GPU ran at about 2,890-2,910 MHz in this workload when running a fixed frame render. The Red Devil ran notably higher, up at 2,970-3,000 MHz instead. That’s around a 60-100 MHz bump in most instances, so the extra power is going somewhere, and it’s sustaining that frequency.

Plotting the Sapphire Pulse as a frame of reference, it holds about 2,900-2,925 MHz at steady state, putting it slightly faster than the Yeston card. The Waifu’s desired frequency just happens to be lower than that of the Pulse and Red Devil.

Gaming Performance

We’ll do a quick look at gaming performance to see how much that extra frequency does or doesn’t matter. This will be a very slimmed-down version of our tests since we only need a few head-to-head results. These were tested on the same drivers, so for purposes of comparing the 9070 XTs to each other, that’s all we need.

FFXIV - 4K

Final Fantasy is first. We chose this to start because the ratio of Waifu-to-non-Waifus in Final Fantasy is extremely disproportionate, and so Yeston should have a natural advantage.

The Red Devil ran at 69.6 FPS AVG in Final Fantasy 14 Dawntrail at 4K, leading the Yeston Atlantis Waifu card by just over 1 FPS AVG. You’d never notice this, but on a technicality, the Red Devil is about 1.9% higher average framerate. The Pulse is about the same as the Atlantis card.

Resident Evil 4 - 1440p

Naturally, because we biased the Final Fantasy test toward the Waifu by nature of the... waifus all over the game, we had to bias the next one toward the Red Devil. Resident Evil 4 is up now, giving PowerColor a distinct and hellish advantage. 

The RX 9070 XT Red Devil draws upon the very evils which it is rendering with its Hellstone to run at 198 FPS AVG, about a 3 FPS lead over the Waifu. That’s about 1.7% ahead. The Waifu outruns the Pulse, though, and you don’t need to be the fastest when you’re running from zombies -- you just need to not be the slowest.

Resident Evil 4 - 4K

The only thing worse than zombies at 1440p is zombies at 4K, so that’s what we’re rendering now. In Resident Evil 4 at 4K, the Red Devil held a 106 FPS AVG, leading the Waifu by about 2 FPS AVG. That’s around 2%.

All of these numbers are expected. We did technically run 4 other games, but they all show the same thing: Typically, we’re in the range of 1-3% improved on the PowerColor card out of the box.

Conclusion

The price for both cards is the same, and they’re both crazy. The good news is that because both are $900, they’re easy to compare.

We have to give both cards credit. Neither are bad. We’ve already reviewed the Yeston RX 9070 XT Sakura Sugar Atlantis OC Edition, in which we also tore down the card. We were fairly positive on it. There were things it could have done better. For example, it doesn’t have as good of a thermal solution as we’ve seen on some of their past cards with this design approach, but it was still fine in pretty much every metric. 

For the Red Devil, the biggest place that card is lacking is in underutilizing the ability to tune its fan curves more. This is something PowerColor seems to consistently have problems with in the cards we test from them. The company could utilize it more to better balance the thermals and acoustics more reasonably.

Overall, though, both cards are pretty close to each other. The Red Devil, in noise-normalized situations, tends to be the better cooler, but the Waifu card, out of the box, tends to be the better cooler. It depends on how you’re going to run them. The Red Devil also has a higher clock out of the box and it actually runs faster. Now, in gaming performance, this difference doesn’t really manifest itself in a big way. 

We did play around with overclocking on both cards. We got a little higher clocks out of the Red Devil. That card also has a higher power budget.

Both cards are too expensive, however. At least both cards do something to try and justify their price.

We’ve compiled a comprehensive timeline of the GPU bans, GPU smuggling, and export controls that impact NVIDIA, AMD, and Intel

The Highlights

• The US blocks exports of advanced GPUs to China to protect national security
• NVIDIA GPUs are highly sought after in China for AI processing
• Our timeline chronicles the US export controls, NVIDIA's responses, and reports of GPU smuggling

This is a comprehensive timeline of the GPU bans, smuggling, and export controls on NVIDIA, AMD, Intel, and other high-tech semiconductor products. We are publishing this as part of our stretch goals for Black Market AI GPU — a viewer-funded film made possible through support on our store, including our new “Blind Eye” T-shirt.
The below timeline accompanies our Black Market AI GPU investigation, our biggest project yet. We spent three weeks in Asia to uncover this story, including two weeks in China and one in Taiwan. We found smugglers, middlemen, and users of so-called “AI” GPUs that the United States government has banned for sale into China.

Credits

Host, Writing, Lead Editing

Steve Burke

Editing

Vitalii Makhnovets
Tim Phetdara

Editing, Graphics

Andrew Coleman

Camera

Tannen Williams

Research and Writing

Ben Benson

We are providing this timeline for free and without third-party ads for our viewers and readers. As this situation has changed frequently and now spans multiple US administrations, we may have missed a few events. However, we believe we have compiled all the major changes – especially since the start of 2025 – that are directly relevant to the story. 

We have attempted to present it as neutrally as feasible and from a place of reporting. We’ve included links to a variety of media and government sources that we believe are appropriate for establishing the timeline of events. We have included statements from NVIDIA in many cases. 

This was a huge team effort at GN and required a massive investment in travel, writing, research, and editing to complete. If you find this information valuable, we ask that you please support us directly by backing our NVIDIA AI GPU Black Market project, buying something from our GN store, or signing up for our Patreon. Thank you.

Timeline

Note on sources: Our intent is to cite primary sources, including government documents, and a variety of secondary sources. In some cases, we link only to secondary news stories. This can occur when we include articles from credible media reports but do not have primary documents to cite. 

2018 GPU Export Controls

August 2018

August 13: The US government created the National Security Commission on Artificial Intelligence (NSCAI) as part of the John S. McCain National Defense Authorization Act for Fiscal Year 2019. The NSCAI had 15 commissioners who were nominated by Congress and the Executive Branch. The NSCAI was tasked with investigating how the United States should compete in AI in the modern age and recommending actions for Congress and the executive branch.

In the words of the original document, the commissioners “shall consider the methods and means necessary to advance the development of artificial intelligence, machine learning, and associated technologies by the United States to comprehensively address the national security and defense needs of the United States.”

• Source:
  • National Defense Bill (NSCAI section starts on page 1963)

2019 GPU Export Controls

May 2019

May 15: Citing national security risks, the US government added Huawei to its Entity List and restricted sales of Huawei’s equipment into the United States. 

• Sources:
  • Federal Register
  • Executive Order
  • Politico: Trump signs order setting stage to ban Huawei from U.S.
  • Reuters: Trump administration hits China's Huawei with one-two punch
  • Reuters: China's Huawei, 70 affiliates placed on U.S. trade blacklist

2020 GPU Export Controls

May 2020

May 19: The United States restricted semiconductor designs, chipsets, and technologies to Huawei and its foreign affiliates. 

• Sources: 
  • Department of Commerce Press Release
  • Federal Register
  • NPR: The Latest U.S. Blow To China's Huawei Could Knock Out Its Global 5G Plans

2021 GPU Export Controls

March 2021

The National Security Commission on Artificial Intelligence (NSCAI) released its final report. The report provided recommendations to “advance the development of artificial intelligence, machine learning, and associated technologies to comprehensively address the national security and defense needs of the United States.”

As part of the report (page 216), the NSCAI recommended the US government and its allies “utilize targeted export controls on high-end semiconductor manufacturing equipment… to protect existing technical advantages and slow the advancement of China’s semiconductor industry.” 

Further on (page 228), the report said, “Looking across the AI stack, the hardware component of the AI stack contains the most viable targets for traditional export controls.” The report (page 231) focused on semiconductor manufacturing equipment for export control rules: “The primary U.S. export control target to constrain competitors’ AI capabilities should be sophisticated semiconductor manufacturing equipment (SME) necessary to manufacture high-end chips.” 

The report mentioned export controls for GPUs (page 500) as a way “to prevent the use of

high-end U.S. AI chips in human rights violations.”

• Source: 
  • NSCAI Final Report – The PDF can be downloaded from here.

April 2021

NSCAI Commissioner Christopher Darby spoke at NVIDIA GTC about the NSCAI’s report to Congress.

• Source: 
  • GTC

October 2021

October 1: The NSCAI officially ended on October 1, 2021. 

• Source: 
  • Executive Government News

2022 GPU Export Controls

August 2022

August 31: NVIDIA filed a Form 8-K with the SEC to inform investors that the US government had immediately blocked exports of its A100 and H100 chips to China, including Hong Kong. The export controls included DGX or other systems that incorporate an A100, H100, or A100X. In the financial documents, NVIDIA said the US government informed it of the export restrictions on August 26, 2022. NVIDIA stated that its third-quarter results included up to $400 million in expected sales to China that were now uncertain due to the export restrictions.

• Sources: 
  • NVIDIA Form 8-K
  • NVIDIA Form 10-Q
  • SEC Filing

September 2022

September 1: NVIDIA filed a new Form 8-K to let customers know that the US government had offered some exemptions for certain chip exports: 

“The U.S. government has authorized exports, reexports, and in-country transfers needed to continue NVIDIA Corporation’s, or the Company’s, development of H100 integrated circuits after the Company filed its Current Report on Form 8-K with the U.S. Securities and Exchange Commission on August 31, 2022. The authorization also allows the Company to perform exports needed to provide support for U.S. customers of A100 through March 1, 2023. Additionally, the U.S. government authorized A100 and H100 order fulfillment and logistics through the Company’s Hong Kong facility through September 1, 2023.”

• Source: 
  • NVIDIA Form 8-K

Following NVIDIA’s SEC filing, media outlets reported the US government ordered NVIDIA to stop selling advanced AI chips to China. 

• NVIDIA statement: 

• “We are working with our customers in China to satisfy their planned or future purchases with alternative products and may seek licenses where replacements aren’t sufficient. The only current products that the new licensing requirement applies to are A100, H100 and systems such as DGX that include them.”– NVIDIA to CNBC

• Sources:
  • CNBC: NVIDIA stock falls after U.S. government restricts chip sales to China
  • The New York Times: U.S. Restricts Sales of Sophisticated Chips to China and Russia
  • Reuters: U.S. officials order NVIDIA to halt sales of top AI chips to China
  • CNN: US orders NVIDIA and AMD to stop selling AI chips to China 
  • Associated Press: China demands US drop tech export curbs after NVIDIA warning

October 2022

October 7: The Department of Commerce’s Bureau of Industry and Security (BIS) implemented a series of export controls to “protect US national security and foreign policy interests.” The new export controls would hinder China’s ability to build high-end semiconductors and purchase advanced chips from the US, including for development of and maintaining supercomputers. 

• Sources: 
  • Department of Commerce Release
  • Federal Register (amended on October 13)

In a briefing with reporters, the US government said the new regulations formalized the guidance previously sent to NVIDIA. The Guardian reported: 

“The new regulations will also severely restrict export of US equipment to Chinese memory chip makers and formalize letters sent to NVIDIA Corp and Advanced Micro Devices Inc (AMD) restricting shipments to China of chips used in supercomputing systems that nations around the world rely on to develop nuclear weapons and other military technologies.”

• Sources: 
  • The Guardian: Biden administration imposes sweeping tech restrictions on China
  • The New York Times: Biden Administration Clamps Down on China’s Access to Chip Technology

November 2022 

November 7: Reuters reported that NVIDIA had created a new AI chip called the A800 GPU for the China market. The A800 would be compliant with US export controls.

• NVIDIA statement:

• “The NVIDIA A800 GPU, which went into production in Q3, is another alternative product to the NVIDIA A100 GPU for customers in China. The A800 meets the US government’s clear test for reduced export control and cannot be programmed to exceed it.” - NVIDIA to Reuters

• Sources:
  • Reuters: Exclusive: NVIDIA offers new advanced chip for China that meets U.S. export controls
  • Tom’s Hardware: NVIDIA Creates New Supercomputer Chip For Chinese Market
  • The Verge: NVIDIA’s selling a nerfed GPU in China to get around export restrictions
  • TechCrunch: NVIDIA touts a slower chip for China to avoid US ban

2023 GPU Export Controls

March 2023

March 21: Reuters reported that NVIDIA had modified the H100 to be compliant with export rules to China.

• NVIDIA’s statements to Reuters:

• “On Tuesday, the company said it has similarly developed a China-export version of its H100 chip. The new chip, called the H800, is being used by the cloud computing units of Chinese technology firms such as Alibaba Group Holding Ltd, Baidu Inc and Tencent Holdings Ltd, a company spokesperson said.” [...]

“The NVIDIA spokesperson declined to say how the China-focused H800 differs from the H100, except that ‘our 800 series products are fully compliant with export control regulations.’” 

• Sources: 
  • Reuters: NVIDIA tweaks flagship H100 chip for export to China as H800
  • Data Center Dynamics: NVIDIA creates pared back H100 GPU for export to China, called H800

June 2023

June 27: The Wall Street Journal reported that the US government is considering expanding export controls for GPUs and AI chips to China. The US Department of Commerce did not comment to the Wall Street Journal. 

• Sources:
  • The Wall Street Journal: U.S. Considers New Curbs on AI Chip Exports to China
  • MarketWatch: White House says it’s focused on being at front end of supply chain for chips, won’t comment on report of possible new ban on exporting AI chips to China
  • Reuters: US mulls new export restriction on computing power in AI chips
  • TechCrunch: China’s AI firms might further lose chip access in new US ban

October 2023

October 17: The US Department of Commerce updated its export compliance for advanced semiconductors and semiconductor manufacturing equipment. The government said: 

“Today’s rules reinforce the October 7, 2022, controls to restrict the PRC’s ability to both purchase and manufacture certain high-end chips critical for military advantage. These updates are necessary to maintain the effectiveness of these controls, close loopholes, and ensure they remain durable.” 

The U.S. government has removed interconnect speed as a criterion for identifying restricted chips. Instead, it will now focus on processor performance and performance density. In a statement, the government said:

“A performance density parameter prevents the workaround of simply purchasing a larger number of smaller datacenter AI chips which, if combined, would be equally powerful as restricted chips.”

• Sources: 
  • Department of Commerce Release
  • Federal Register
  • Bureau of Industry and Security Document on Performance Density (page  21-22)
  • The Register: Biden has brought the ban hammer down on US export of AI chips to China
  • Center for Strategic & Internal Studies (posted on Oct. 18)

As part of the announcement, the administration told reporters the new restrictions affect NVIDIA’s A800 and H800 chips. A few days prior, Reuters reported that the administration would soon announce new export rules. 

• Sources:
  • Reuters: Exclusive: US tackles loopholes in curbs on AI chip exports to China
  • CNBC: U.S. curbs export of more AI chips, including NVIDIA H800, to China
  • Reuters: NVIDIA details advanced AI chips blocked by new export controls
  • Tom’s Hardware: US Prohibits Exports of NVIDIA’s A800 and H800 to China, Blacklists Chinese GPU Developers
  • The Verge: NVIDIA’s H800 AI chip for China is blocked by new export rules

October 23: NVIDIA filed a Form 8-K with the SEC that said the new export rules impact its A100, A800, H100, H800 and L40S chips. NVIDIA said it “does not anticipate that the accelerated timing of the licensing requirements will have a near-term meaningful

impact on its financial results.” 

• Sources: 
  • NVIDIA Form 8-K
  • BBC: US orders immediate halt to some AI chip exports to China, NVIDIA says
  • Tom’s Hardware: US Govt Speeds Up Export Restrictions for NVIDIA’s GPUs

December 2023

December 6: In a meeting with reporters in Singapore, NVIDIA said that it was working on new chips that comply with the government’s rules. 

• NVIDIA statement:

• “NVIDIA has been working very closely with the U.S. government to create products that comply with its regulations. Our plan now is to continue to work with the government to come up with a new set of products that comply with the new regulations that have certain limits.” – NVIDIA CEO Jensen Huang, as reported in Reuters

• Sources:
• Reuters: NVIDIA working closely with US to ensure new chips for China are compliant with curbs
  • Reposted in CNBC

December 28: NVIDIA released a new version of RTX 4090 for the China market. The new chip, called the GeForce RTX 4090D, would be compliant with US export control restrictions. 

• NVIDIA statements: 

• “The GeForce RTX 4090 D has been designed to fully comply with U.S. government export controls. While developing this product, we extensively engaged with the U.S. government.” - NVIDIA to Reuters

• “In 4K gaming with ray tracing and deep-learning super sampling (DLSS), the GeForce RTX 4090D is about five percent slower than the GeForce RTX 4090 and it operates like every other GeForce GPU, which can be overclocked by end users.” – NVIDIA to The Register

• Sources: 
  • Reuters: NVIDIA launches new gaming chip for China to comply with US export controls
  • The Register: NVIDIA slowed RTX 4090 GPU by 11 percent, to make it 100 percent legal for export to China
  • The Verge: NVIDIA is releasing a slower RTX 4090 in China to comply with US restrictions

2024 GPU Export Controls

February 2024

February 1: Reuters reported that NVIDIA had prepared new GPUs for China, including the H20. Several sources told Reuters that the new offerings are less powerful than similar chips from Huawei. 

• Sources:
  • Reuters: Exclusive: NVIDIA’s new China-focused AI chip set to be sold at similar price to Huawei product
  • Tom’s Hardware: New NVIDIA AI GPUs designed to get around U.S. export bans come to China — H20, L20, and L2 to fill void left by restricted models

July 2024

July 22: Reuters reported that NVIDIA is creating a new GPU for the China market based on its Blackwell chips. Sources told Reuters that the chip would be a version of the Blackwell B200. NVIDIA did not publicly disclose the specifications. 

• Sources:
  • Reuters: Exclusive: NVIDIA preparing version of new flagship AI chip for Chinese market
  • Tom’s Hardware: NVIDIA preparing a China-focused variant of its B200 Blackwell AI GPU to comply with US export regulations
  • HPCWire: NVIDIA Prepares New AI Chip for China Amid Ongoing US Export Controls

December 2024

December 2: The US government expanded rules that limit the export of high memory bandwidth (HBM) and advanced semiconductor equipment. 

The Center for Strategic & International Studies explained the new rules on HBM: 

“The December 2024 controls change that by adopting for the first time country-wide restrictions on the export of advanced HBM to China as well as an end-use and end-user controls on the sale of even less advanced versions of HBM. The goal of these controls is, unsurprisingly, to degrade China’s AI industry.” [...]

“Modern AI chips not only require a lot of memory capacity but also an extraordinary amount of memory bandwidth. Bandwidth refers to the amount of data a computer’s memory can transfer to the processor (or other components) in a given amount of time. With low-bandwidth memory, the processing power of the AI chip often sits around doing nothing while it waits for the necessary data to be retrieved from (or stored in) memory and brought to the processor’s computing resources.”

• Sources:
  • Department of Commerce Document
  • Federal Register 
  • Government Presentation
  • Center for Strategic & International Studies: Understanding the Biden Administration’s Updated Export Controls

2025 GPU Export Controls

January 2025

January 13: The US government tightened its export controls by introducing national chip caps for many countries, except for 18 allies. The new restrictions would be called the AI Diffusion Rule. The rule would go into effect in May 2025.  

• NVIDIA statement:

• “It makes no sense for the Biden White House to control everyday datacenter computers and technology that is already in gaming PCs worldwide, disguised as an anti-China move. The extreme ‘country cap’ policy will affect mainstream computers in countries around the world, doing nothing to promote national security but rather pushing the world to alternative technologies. AI is mainstream computing – ubiquitous and essential as electricity. This last-minute Biden Administration policy would be a legacy that will be criticized by U.S. industry and the global community. We would encourage President Biden to not preempt incoming President Trump by enacting a policy that will only harm the U.S. economy, set America back, and play into the hands of U.S. adversaries.” – Ned Finkle, Vice President of Government Affairs, NVIDIA, to Bloomberg (Twitter link)

• Sources: 
  • US Government Fact Sheet
  • Federal Register
  • Federal Register
  • Bloomberg: Biden to Further Limit NVIDIA AI Chip Exports in Final Push
  • Tom’s Hardware: NVIDIA and SIA fire back at US gov's new export restrictions on AI GPUs to China

February 2025

February 26: NVIDIA filed its 10-K annual report with the SEC. In the 10-K, NVIDIA revealed that Singapore was the second-largest geographical source of revenue in 2024, behind the United States. Taiwan was third, and China was fourth.

Within the report, NVIDIA said: 

“Singapore represented 18% of fiscal year 2025 total revenue based upon customer billing location. Customers use Singapore to centralize invoicing while our products are almost always shipped elsewhere. Shipments to Singapore were less than 2% of fiscal year 2025 total revenue.”

• Source:
  • NVIDIA 10-K

February 27: Speculation began about AI GPUs being smuggled from Singapore to China. In late February, authorities in Singapore arrested three people for fraud involving servers that may contain AI GPUs. Singapore’s government granted the three people bail a few weeks later.

NVIDIA declined to comment to CNBC. 

• Sources:
  • ChannelNewsAsia: 3 men charged with fraud, cases linked to alleged movement of Nvidia chips
  • CNBC: NVIDIA’s unofficial exports to China face scrutiny after arrest of silicon smugglers in Singapore
  • Tom’s Hardware: Singapore police bust major ring smuggling NVIDIA GPUs to China-based DeepSeek: Report
  • TechCrunch: Singapore grants bail for NVIDIA chip smugglers in alleged $390M fraud

April 2025

April 9: NPR reported that the US government would not add export controls for the H20 chip after NVIDIA CEO Jensen Huang attended a dinner at Mar-A-Lago. The dinner reportedly cost $1 million per head. The outlet said it was unclear whether Jensen Huang met with US President Trump directly. 

NVIDIA declined to comment to NPR.

• Sources:
  • NPR: Trump administration backs off NVIDIA H20 chip crackdown after Mar-a-Lago dinner

April 15: In a SEC filing, NVIDIA said the US government sent the company new export rules on April 9. According to NVIDIA, the H20 and all chips with the H20’s memory bandwidth or interconnect bandwidth will now need licenses to export to China. NVIDIA said the new rules would cost the company $5.5 billion in charges due to current H20 chip inventory and prior sales. NVIDIA declined to comment further to the BBC. 

• Sources: 
  • NVIDIA Form 8-K
  • BBC: NVIDIA shares plunge amid $5.5bn hit over export rules to China
  • NPR: NVIDIA discloses that U.S. will limit sales of advanced chips to China after all
  • Reuters: US issues export licensing requirements for NVIDIA, AMD chips to China

April 16: The US government released an investigative report on DeepSeek and requested information from NVIDIA about its AI GPUs. Through a letter sent to NVIDIA CEO Jensen Huang, the US government asked NVIDIA for a list of its customers in China and many countries in Asia, including Singapore. The government requested all communication between NVIDIA and DeepSeek. 

• Sources:
  • US government press release
  • DeepSeek report
  • Government letter to Jensen Huang

The US Department of Commerce confirmed that it has issued new export control rules for AI chips. The Commerce Department provided a statement to The Wall Street Journal: 

“The Commerce Department is issuing new export licensing requirements on the NVIDIA H20, AMD MI308, and their equivalents.”

• Sources:
  • The Wall Street Journal: U.S. Confirms New Export Curbs on NVIDIA and AMD Chips

April 28: The Wall Street Journal reported that Huawei is expected to release its new AI chip, the Ascend 910D, soon. According to the Wall Street Journal’s sources, Huawei expects the Ascend 910D to be about as powerful as an NVIDIA H100. 

• Sources:
  • The Wall Street Journal: China’s Huawei Develops New AI Chip, Seeking to Match NVIDIA 
  • NetworkWorld: Huawei steps up AI chip race with Ascend 910D, targeting NVIDIA’s high ground

April 30: Anthropic, an AI startup backed by Amazon, called on the US government to increase export control restrictions to China. As part of a blog post, Anthropic said the government needs to improve its export enforcement to reduce smuggling. The company cited examples of chips being smuggled with “prosthetic baby bumps” and “live lobsters.”

In a response, NVIDIA said: 

“American firms should focus on innovation and rise to the challenge, rather than tell tall tales that large, heavy, and sensitive electronics are somehow smuggled in ‘baby bumps’ or ‘alongside live lobsters.’” – NVIDIA to CNBC

• Sources:
  • Anthropic blog post
  • CNBC: NVIDIA says Anthropic is telling ‘tall tales’ in its defense of U.S. AI chip restrictions on China
  • Tom’s Hardware: Despite NVIDIA claims, Chinese smugglers have used live lobsters and fake baby bumps to traffic chips
  • Hong Kong Customs release
  • China Customs release 

May 2025

May 1: Jensen Huang spoke with the House Foreign Affairs Committee to discuss domestic manufacturing and the importance of AI. NVIDIA posted the remarks online.

• Sources:
  • NVIDIA Newsroom Twitter Post
  • Tom’s Hardware: NVIDIA warns U.S. AI hardware export rules could backfire, empowering Huawei to define global standards

May 7: Following a report in Bloomberg, the US Department of Commerce confirmed that it will not implement the AI Diffusion Rule that was created during the prior administration. The rule would have gone into effect on May 15, 2025. 

The Department of Commerce released a statement to CNBC: 

“The Biden AI rule is overly complex, overly bureaucratic, and would stymie American innovation. We will be replacing it with a much simpler rule that unleashes American innovation and ensures American AI dominance.” 

NVIDIA released a statement: 

“We welcome the Administration’s leadership and new direction on AI policy. With the AI Diffusion Rule revoked, America will have a once-in-a-generation opportunity to lead the next industrial revolution and create high-paying U.S. jobs, build new U.S.-supplied infrastructure, and alleviate the trade deficit.”

• Sources: 
  • Bloomberg: Trump to Rescind Global Chip Curbs, Prep New AI Restrictions
  • Tom’s Hardware: NVIDIA celebrates dumping of Biden-era AI chip export rules — simpler new policy promised
  • CNBC: Trump administration set to end Biden’s U.S. chip export restrictions
  • NVIDIA Twitter Account

May 9: Reuters reported that NVIDIA is preparing a cut down version of the H20 for the Chinese market. Reuters sources said the chip would be ready in July. NVIDIA declined to comment. 

• Sources: 
  • Reuters: Exclusive: NVIDIA modifies H20 chip for China to overcome US export controls, sources say
  • Tom’s Hardware: NVIDIA readies cut-down HGX H20 GPU for China to comply with export control rules

May 13: The US government formally rescinded the previous administration’s AI Diffusion Rule, which was announced in January 2025. The government also announced actions to strengthen export controls for AI chips, including restrictions on using several Huawei Ascend chips

NVIDIA declined to comment to CNBC on the new export restrictions. 

• Sources:
  • US government press release 
  • BIS policy statement 
  • US government guidance on using Huawei Ascend chips
  • The Register: Trump ends Biden-era dream to cap US AI chip exports
  • Bloomberg: US Warns That Using Huawei AI Chip ‘Anywhere’ Breaks Its Rules
  • Tom’s Hardware: U.S. issues worldwide crackdown on using Huawei Ascend chips, says it violates export controls
  • CNBC: Trump administration’s next wave of China AI chip export rules are yet another obstacle for NVIDIA 

May 15: A bipartisan group of legislators introduced the Chip Security Act that is intended to stop smuggling of high-end AI chips. 

The Hill summarized the proposed legislation: “The legislation, titled the Chips Security Act, would require companies to ensure the location-verification abilities of their high-end AI chips and to report when a product has been diverted or changed location. It follows recent reports of increased smuggling of chips, including those made by NVIDIA, into China despite tight export controls.”

NVIDIA declined to comment to The Register.

• Sources: 
  • Chip Security Act text
  • The Hill: Bipartisan House lawmakers propose bill to ‘stop smuggling’ of AI chips
  • Reuters: U.S. lawmakers introduce bill to address AI chip smuggling
  • The Register: Plan to keep advanced chips from China with tracking tech gains support in Congress

May 16: The Financial Times reported that NVIDIA intends to create a research and design center in Shanghai. 

• NVIDIA statement: 

• “We are not sending any GPU designs to China to be modified to comply with export controls.” - NVIDIA to CNBC 

• Sources: 
  • Financial Times: NVIDIA plans Shanghai research centre in new commitment to China
  • CNBC: NVIDIA says it is not sending GPU designs to China after reports of new Shanghai operation

May 19: NVIDIA CEO Jensen Huang told Bloomberg in a TV interview that he didn’t see any “evidence” of any AI chip diversion. Tom’s Hardware summarized Jensen Huang’s quote: 

“Governments understand that diversion is not allowed, and there’s no evidence of any AI chip diversion — recognize our data center GPUs are massive; these are massive systems. The Grace Blackwell system is nearly two tons, and so you’re not going to be shipping — you’re not going to be putting that in your pocket or your backpack anytime soon. And so, these systems are fairly easy to keep track of... but the important thing is that the countries and the companies that we sell to recognize that diversion is not allowed, and everybody would like to continue to buy NVIDIA technology, and so they very well monitor themselves very carefully and they’re quite careful about that.”

• Sources:
  • Bloomberg: NVIDIA CEO Says ‘No Evidence of Any AI Chip Diversion’
  • Tom’s Hardware: NVIDIA CEO Jensen Huang says ‘There’s no evidence of any AI chip diversion’

May 21: At Computex 2025, NVIDIA CEO Jensen Huang called the US export controls a “failure.” He said that NVIDIA’s market share in China has dropped from 95% to 50% due to the restrictions. The Guardian quoted Jensen Huang as saying: 

“The local companies are very, very talented and very determined, and the export control gave them the spirit, the energy and the government support to accelerate their development.” [...]

“I think, all in all, the export control was a failure.” [...]

“China has a vibrant technology ecosystem, and it’s very important to realise that China has 50% of the world’s AI researchers, and China is incredibly good at software.”

• Sources: 
  • CNBC: Jensen Huang says U.S. chip restrictions have cut NVIDIA’s China market share nearly in half
  • The Guardian: US chip export controls are a ‘failure’ because they spur Chinese development, NVIDIA boss says
  • The New York Times: NVIDIA’s Chief Says U.S. Chip Controls on China Have Backfired

May 27: Reuters reported that NVIDIA plans to launch a new, cheaper Blackwell-based GPU for the China market to comply with US export rules. 

• NVIDIA statement: 

• "Until we settle on a new product design and receive approval from the U.S. government, we are effectively foreclosed from China's $50 billion data center market." - NVIDIA to Reuters

• Sources: 
  • Reuters: Exclusive: NVIDIA to launch cheaper Blackwell AI chip for China after US export curbs, sources say
  • SiliconANGLE: Report: NVIDIA racing to develop new, scaled-down Blackwell GPUs for China

May 28: During NVIDIA’s quarterly earnings, CEO Jensen Huang said the company was writing off unsold H20 inventory due to export controls. VentureBeat posted Jensen Huang’s quote from earnings: 

“Let me share my perspective on some topics we’re frequently asked on export control. China is one of the world’s largest AI markets and a springboard to global success with half of the world’s AI researchers based there. The platform that wins China is positioned to lead globally today. However, the $50 billion China market is effectively closed to U.S. industry. The H20 export ban ended our Hopper data center business in China. We cannot produce Hopper further to comply. As a result, we are taking a multibillion-dollar write-off on inventory that cannot be sold or repurposed. We are exploring limited ways to compete, but hopper is no longer an option.”

• Sources:
  • VentureBeat: NVIDIA CEO takes a shot at U.S. policy cutting off AI chip sales to China
  • PC Gamer: NVIDIA’s Hopper GPUs are now dead to the Chinese market after export controls that made the company take a 'multibillion-dollar write-off'

June 2025

June 12: NVIDIA’s CEO Jensen Huang told CNN the company will no longer include sales and revenue from China in its forecasts. In a response to a question from CNN about whether the US government would lift its export controls, Jensen Huang said: 

“I’m not counting on it but, if it happens, then it will be a great bonus. I’ve told all of our investors and shareholders that, going forward, our forecasts will not include the China market.” 

• Sources:
  • CNN: NVIDIA will stop including China in its forecasts amid US chip export controls, CEO says

June 18: Several media reported on rumors about NVIDIA preparing to launch a “RTX 5090 DD” for the China market. The new card would allegedly reduce the memory specifications compared to the RTX 5090D. 

• Sources: 
  • Tom’s Hardware: NVIDIA planning new RTX 5090 'DD' variant for China — 24GB card with tweaked GPU latest attempt to comply with strict export restrictions
  • WCCFTech: NVIDIA Preps GeForce RTX 5090 DD For China As Export-Compliant Model, Reportedly Features Blackwell GB202-240 GPU

June 23: Reuters reported that DeepSeek is supporting China’s military and intelligence operations, based on an interview with a senior US State Department official. The official said DeepSeek was using “shell companies” in Southeast Asia to circumvent export restrictions. 

Reuters included comments from NVIDIA: 

“‘We do not support parties that have violated U.S. export controls or are on the U.S. entity lists,’ an NVIDIA spokesman said in a prepared statement, adding that ‘with the current export controls, we are effectively out of the China data center market, which is now served only by competitors such as Huawei.’” [...]

“‘Our review indicates that DeepSeek used lawfully acquired H800 products, not H100,’ an NVIDIA spokesman said, responding to a Reuters query about DeepSeek's alleged usage of H100 chips.” 

DeepSeek did not respond to an inquiry from Reuters. 

• Sources:
  • Reuters: Exclusive: DeepSeek aids China's military and evaded export controls, US official says
  • Asia Times: DeepSeek gets NVIDIA’s high-end GPUs via Singapore: US official

June 26: The Information reported that DeepSeek’s next AI model has been delayed due to a shortage of NVIDIA AI GPUs in China. 

• Sources:
  • The Information: DeepSeek’s Progress Stalled by U.S. Export Controls
  • Tom’s Hardware: AI disruptor DeepSeek's next-gen model delayed by NVIDIA GPU export restrictions to China — short supply of AI GPUs hinders development

July 2025

July 4: Bloomberg reported the US Department of Commerce is preparing a new export controls rule that would restrict the export of AI chips to Malaysia and Thailand. The rule’s goal would be to reduce AI chip smuggling to China. Based on its sources, Bloomberg said the export controls rule had not yet been finalized.  

• Sources: 
  • Bloomberg: US Plans AI Chip Curbs on Malaysia, Thailand Over China Concerns

July 10: Bloomberg reported that NVIDIA CEO Jensen Huang met with US President Donald Trump at the White House before traveling overseas to China. NVIDIA did not immediately respond to a request for comment from Reuters. 

• Sources: 
  • Bloomberg: NVIDIA’s Jensen Huang Meets with Trump Ahead of CEO’s China Trip
  • Reuters: NVIDIA CEO Huang to meet Trump before China trip, source says
  • CNBC: Trump hosts Jensen Huang at White House as NVIDIA tops $4 trillion market cap

July 11: In a public letter, a bipartisan group of US senators requested NVIDIA CEO Jensen Huang to avoid meeting with Chinese companies in an upcoming China trip that violate US laws or develop military applications that could undermine national security. 

Reuters included a response from NVIDIA about the senators’ letter: 

“An NVIDIA spokesperson said, ‘American wins’ when its technology sets ‘the global standard,’ and that China has one of the largest bodies of software developers in the world. AI software ‘should run best on the U.S. technology stack, encouraging nations worldwide to choose America,’ the spokesperson said.” 

• Sources: 
  • United States Senate letter to Jensen Huang
  • The Hill: Bipartisan senators press NVIDIA CEO over China trip
  • Reuters: US senators warn NVIDIA CEO about upcoming China trip
  • International Business Times: NVIDIA CEO Huang to Face Chinese Officials Over AI Export Curbs Just as Company Touches $4 Trillion Milestone

July 14: NVIDIA said it would soon resume sales of the H20 for customers in China. NVIDIA provided the following update in a blog post: 

“[Jensen] Huang also provided an update to customers, noting that NVIDIA is filing applications to sell the NVIDIA H20 GPU again. The U.S. government has assured NVIDIA that licenses will be granted, and NVIDIA hopes to start deliveries soon.”

The White House did not respond to a request for comment from CNN.

• Sources: 
  • NVIDIA Blog: NVIDIA CEO Jensen Huang Promotes AI in Washington, DC and China
  • CNN: NVIDIA says it will restart sales of a key AI chip to China, in a reversal of US restrictions
  • Reuters: Chinese firms rush to buy NVIDIA AI chips as sales set to resume
  • Associated Press: NVIDIA to resume sales of highly desired AI computer chips to China

The Malaysian government began requiring trade permits for all high-performance AI chips acquired from the United States. In a statement, the Malaysian government said: 

“The Ministry of Investment, Trade and Industry (MITI) would like to announce that, effective immediately, all exports, tranships and transits of high-performance AI chips of US origin are subject to a Strategic Trade Permit. These powers are provided for under Section 12 of the Strategic Trade Act 2010 (STA 2010), a Catch-All Control provision which requires individuals or companies to notify the relevant authority at least 30 days before exporting, transhipping, or bringing in transit any item not expressly listed in the Strategic Items List (SIL), if the individual or company knows or have reasonable grounds to suspect the item will be misused, or used for a restricted activity.

This initiative serves to close regulatory gaps while Malaysia undertakes further review on the inclusion of high-performance AI chips of US origin into the SIL of the STA 2010. Malaysia stands firm against any attempt to circumvent export controls or engage in illicit trade activities by any individual or company, who will face strict legal action if found violating the STA 2010 or related laws.”

• Sources: 
  • Malaysia Government Release
  • Reuters: Malaysia says trade permit required for AI chips of U.S. origin
  • Bloomberg: Malaysia Controls AI Chip Exports As US Targets China Smuggling

July 15: DigiTimes reported that NVIDIA is preparing a new AI GPU for the China market, the RTX 6000D. DigiTimes claimed the card would become available in the third quarter of 2025, according to its sources in the supply chain. 

• Sources: 
  • DigiTimes: Exclusive: Jensen Huang's third visit to China in 2025; RTX 6000D aims for two million shipments
  • Tom’s Hardware: NVIDIA reportedly preparing RTX 6000D for Chinese market to comply with U.S. export controls — fabricated on TSMC N4, featuring GDDR7 memory capable of delivering 1,100 GB/s of bidirectional bandwidth

July 24: The Financial Times reported that more than $1B worth of NVIDIA’s AI chips had been smuggled to China. In response, NVIDIA said that building datacenters with “smuggled products” was a “losing proposition.” 

• NVIDIA statement: 

• “Trying to cobble together datacenters from smuggled products is a losing proposition, both technically and economically. Datacenters require service and support, which we provide only to authorized NVIDIA products.” - NVIDIA to CNBC

• Sources:
  • Financial Times: NVIDIA AI chips worth $1bn smuggled to China after Trump export controls
  • CNBC: NVIDIA addresses AI chip smuggling, says bootleg data centers are a ‘losing proposition’

July 28: The Financial Times reported that the US Commerce Department was not going to make “tough moves” to tighten export controls to China. According to the report, the US government would try to secure a better trade deal with China ahead of negotiations in Stockholm.

The Washington Post reported that several congressional members had warned the US administration against loosening its export controls for AI GPUs. NVIDIA and the US Commerce Department did reply to requests for comment to The Washington Post.

Several national security experts voiced their concern by sending a letter to the US Commerce Department. 

• Sources: 
  • Financial Times: Donald Trump freezes export controls to secure trade deal with China
  • The Washington Post: Trump’s retreat on China chip ban triggers policy spat
  • Tom’s Hardware: Trump freeze on export restrictions to China reportedly in aid of trade talks — White House seeking face-to-face with Xi Jinping as dissenters warn H20 reversal is a dangerous mis-step

July 29: Reuters reported that NVIDIA had ordered 300,000 more H20 chips from TSMC due to strong demand from its customers in China. Several weeks prior, NVIDIA said it would resume sales of the H20 chip to China.

NVIDIA declined to comment to Reuters.

• Sources: 
  • Reuters: Exclusive: NVIDIA orders 300,000 H20 chips from TSMC due to robust China demand, sources say
    • Repost in MSN
• Hot Hardware: TSMC Secures 300,000 H20 AI Chip Order As NVIDIA Boosts Supply To China

July 31: The New York Times reported that Chinese government officials asked NVIDIA for information about security risks associated with its H20 chip. NVIDIA denied having “backdoors” in its chips. 

• NVIDIA statement: 

• “Cybersecurity is critically important to us. NVIDIA does not have ‘backdoors’ in our chips that would give anyone a remote way to access or control them.” - NVIDIA to CNBC

• Sources: 

• The New York Times: China Summons NVIDIA Over ‘Backdoor Security’ Risks of A.I. Chips
• CNBC: NVIDIA denies its China-bound H20 AI chips have ‘backdoors’ after Beijing’s security concerns
• Reuters: NVIDIA says its chips have no 'backdoors' after China flags H20 security concerns
• Associated Press: China summons NVIDIA over ‘backdoor safety risks’ in H20 chips

August 2025

August 4: A government official told Bloomberg the United States is exploring adding location trackers for AI chips. Bloomberg quoted the official as saying, “There is discussion about potentially the types of software or physical changes you could make to the chips themselves to do better location-tracking.”

• Sources: 
  • Bloomberg: US Explores Location Trackers for AI Chips, Official Says
  • The Register: Uncle Sam floats tracking tech to keep AI chips out of China

August 5: In a blog post, NVIDIA said that its GPU products do not have backdoors or kill switches. 

• Sources: 
  • NVIDIA Blog: No Backdoors. No Kill Switches. No Spyware.
  • CNBC: NVIDIA says its AI chips don’t have a ‘kill switch’ after Chinese accusation

The US Department of Justice announced it had arrested two people in California for smuggling high-end GPUs to China that purportedly amount to “tens of millions of dollars’ worth of sensitive microchips used in artificial intelligence (AI) applications.” The BBC reported that court documents say the shipments included the NVIDIA H100 and RTX 4090. 

• NVIDIA statement:

• “This case demonstrates that smuggling is a nonstarter. We primarily sell our products to well-known partners, including OEMs, who help us ensure that all sales comply with U.S. export control rules. Even relatively small exporters and shipments are subject to thorough review and scrutiny, and any diverted products would have no service, support, or updates.” - NVIDIA to TechCrunch

• Sources: 
  • Department of Justice release
  • Bloomberg: US Charges Chinese Nationals With NVIDIA Chips Export Breach
  • Reuters: Two Chinese nationals in California accused of illegally shipping NVIDIA AI chips to China
  • New York Post: Chinese nationals living in US charged with smuggling millions worth of NVIDIA’s powerful AI chips to Beijing
  • TechCrunch: Two arrested for smuggling AI chips to China; NVIDIA says no to kill switches
  • BBC: Chinese nationals charged with exporting NVIDIA AI chips to China

August 10: The Financial Times reported that NVIDIA would give the US government 15% of its revenue from H20 chip sales from customers in China. The deal is reportedly part of an agreement that would allow NVIDIA to acquire export licenses from the Commerce Department in order to sell the H20 chip in China. AMD would be subject to the same rules for the MI308. 

• NVIDIA statement: 

• “We follow rules the U.S. government sets for our participation in worldwide markets. While we haven’t shipped H20 to China for months, we hope export control rules will let America compete in China and worldwide. America cannot repeat 5G and lose telecommunication leadership. America’s AI tech stack can be the world’s standard if we race.” - NVIDIA to the Associated Press

• Sources:
  • Financial Times: NVIDIA and AMD to pay 15% of China chip sale revenues to US government
  • The Associated Press: NVIDIA and AMD to pay 15% of China chip sale revenue to US government in an unusual agreement
  • The New York Times: U.S. Government to Take Cut of NVIDIA and AMD A.I. Chip Sales to China
  • BBC: NVIDIA and AMD to pay 15% of China chip sales to US

August 11: According to a report in Bloomberg, US President Trump said he was open to allowing NVIDIA to sell a modified Blackwell chip for the China market. The US President also said that he has negotiated with NVIDIA CEO Jensen Huang about the deal to allow H20 sales in China. 

• Sources:
  • Bloomberg: Trump Open to NVIDIA Selling Scaled-Back Blackwell Chip to China
  • The Register: Trump seeing green as he weighs deal to allow NVIDIA Blackwell GPU sales to China
  • NPR: Trump says NVIDIA will hand the U.S. 15% of its H20 chip sales to China
  • Reuters: Trump opens door to sales of version of NVIDIA’s next-gen AI chips in China
    • Repost in MSN

August 12: Bloomberg reported that Chinese officials had “urged local companies” to avoid purchasing and using NVIDIA’s H20 chip, especially for national security and government work. According to Bloomberg, China questioned companies whether they had found security problems with NVIDIA’s chips.  

Bloomberg included commentary from NVIDIA: 

“AMD declined to comment, while NVIDIA said in a statement that ‘the H20 is not a military product or for government infrastructure.’ China has ample supplies of domestic chips, NVIDIA said, and ‘won’t and never has relied on American chips for government operations.’

• Sources:
  • Bloomberg: China Urges Firms Not to Use NVIDIA H20 Chips In New Guidance
  • Reuters: China cautions tech firms over NVIDIA H20 AI chip purchases, sources say

CNBC reported that the Trump Administration was still working on the details for how to implement the 15% export tax on NVIDIA and AMD for selling certain chips to China. 

• NVIDIA statement:

• “We follow rules the U.S. government sets for our participation in worldwide markets.” - NVIDIA to CNBC

• Sources:
  • CNBC: White House says it’s working out legality of NVIDIA and AMD China chip deals
  • Tom’s Hardware: White House confirms it's still figuring out the legality of the revenue-sharing NVIDIA and AMD deal for China GPU sales — 'The legality of it, the mechanics of it, is still being ironed out'

August 13: Reuters reported that US officials have covertly placed “location-tracking devices” in targeted shipments with advanced chips in an effort to catch chip smuggling to China. Unnamed sources told Reuters that the tracking devices had been placed in shipments of OEM servers, including from Dell and Supermicro. 

NVIDIA declined to comment to Reuters. 

• Sources: 
  • Reuters: Exclusive: US embeds trackers in AI chip shipments to catch diversions to China, sources say
  • Tom’s Hardware: U.S. authorities allegedly placed secret tracking devices in AI chip shipments to China — report claims targeted shipments from Dell and Super Micro containing NVIDIA and AMD chips had trackers in packaging and servers themselves

TweakTown: US authorities secretly place location tracking devices in targeted AI chip shipments to China

Our review of the AMD RX 9060 XT 16GB tests the card against the NVIDIA RTX 5060, RTX 5060 Ti 16GB, RX 7700 XT, 6600 XT, and similar GPUs

The Highlights

• The RX 9060 XT comes with either 8GB or 16GB of VRAM
• The 8GB variant has VRAM limitation issues
• AMD, like NVIDIA, is taking advantage of consumers more aggressively than we've seen in the past
• Original MSRP: $350 (16GB) and $300 (8GB)
• Release Date: June 5, 2025

Table of Contents

• AutoTOC

Intro

Quick version up front: Against the RTX 5060 (read our review), the RX 9060 XT 16GB is anywhere from 2% to 17% better at 1080p rasterized, typically 10%, the same to 22% better at 1440p, and worse to 21% better at 4K, with more variability here. In RT, the 9060 XT again ranges from significantly worse than the RTX 5060 in Black Myth: Wukong to can’t-divide-by-zero better in some specific tests where the 5060 ran out of VRAM. This would also happen to the 9060 XT 8GB model, which…exists. Against itself, AMD’s ray tracing performance has continued to show significant improvement over the last generation, with it sometimes equating the 7700 XT (read our review) in RT scenarios.
Now that we have the important stuff out of the way, before we get started, a quick callback to a few months ago.

Editor's note: This was originally published on June 4, 2025 as a video. This content has been adapted to written format for this article and is unchanged from the original publication.

Credits

Test Lead, Host, Writing, Editing

Steve Burke

Testing

Patrick Lathan

Testing, Editing

Mike Gaglione

Camera, Video Editing

Tim Phetdara

Camera

Andrew Coleman

Camera, Testing

Tannen Williams

Writing, Web Editing

Jimmy Thang

Unfortunately, we’re back to that because AMD can’t just launch a product and shut the f*ck up. It has to shove its foot so deep into its own mouth that it comes out the other side. 

Today, we’re reviewing AMD’s RX 9060 XT. The 9060 XT comes in 8GB and 16GB options, but beyond this, it comes with the usual side of AMD executive tweets.

AMD’s Frank Azor and guy who once bet Twitter $10 that he’d have better supply than NVIDIA during COVID said this:

“Majority of gamers are still playing at 1080p and have no use for more than 8GB of memory. Most played games WW are mostly esports games. We wouldn’t build it if there wasn’t a market for it. If 8GB isn’t right for you then there’s 16GB. Same GPU, no compromise, just memory options.”

First of all, AMD is the king of building things that there’s no market for. For example, these include:

• The 2004 Personal Internet Communicator it abandoned in under 2 years
• Bulldozer
• Anti-Lag+ the first time, before pulling it because it got people VAC banned
• The FX-9370
• The Radeon Pro Duo with almost no actual market
• Athlon 64 X2 3800+ EE SFF
• Radeon VII
• Vega Frontier Edition
• Vega 64
• AMD RX 6300
• The Radeon 6500 XT with PCIe Gen 4 x 4
• The HD 2900 XT
• AMD Fusion Render Cloud in 2009
• AMD Live!
• And finally, the AMD mountain bike, which is still the proudest we’ve ever been of a product that GN killed entirely on its own.

Secondly, Azor wrote, “same GPU, no compromise” while having half the VRAM is literally a compromise. That is the definition of a compromise in the same way Corsair’s i500 was such a compromise as an SFF PC that it removed all the “no compromise” language after our review. Rebranding “compromise” as “options” is the type of thing you put on your job application to become a Chief Architect of Marketing. 

Oh… OK, bad example.

In June 2020, AMD posted a blog post saying that 4GB of VRAM “is evidently not enough for today’s games.” In January 2022, AMD hides the blog post just ahead of the launch of the RX 5500 XT 4GB GPU, then panicked when it was discovered and unhid the blog post.

In 2020 after NVIDIA’s paid 8K marketing campaign, Azor tweeted “I guess we’ll scratch our 16K gaming message” in a jab at NVIDIA. 

In November 2020, AMD partners with Sony to slap a giant “8K” sticker on the PlayStation 5 boxes that it later removed the label from.

Also in November, 2020, Frank Azor tweeted about successfully buying his own product in a market with a massive GPU shortage presumably to save $10 from a bet.

In 2017, AMD posted a video that said “Poor Volta” before having its ass handed to it by NVIDIA for the next several generations.

Also in years past, the Fury is called the “overclocker’s dream.”

In 2016, AMD former Corporate Vice President Roy Taylor sat next to us at the RX 480 (watch our review) launch event and looked at us when we were looking up the Fury X (read our review) specs. He asked why we needed them. We said because we were trying to remember if it was 4GB of memory, and he said, “we’ll never make that mistake again.”

That’s the same launch event that had two RX 480s in CrossFire to show them beating a single GTX 1080 in Ashes of the Singularity, even though AMD’s marketing team ran the CPU benchmark for one brand and the GPU benchmark for the other in a completely incomparable benchmark.

AMD is single-handedly keeping Podithodontists employed by requiring so many emergency foot-from-mouth extraction surgeries.

The point is that an 8GB card should not exist for $300 today. Not for NVIDIA, not for AMD. 8GB cards do have a place -- he’s right about that. But it’s not in $300 solutions. We had 8GB cards for $200 years ago when the RX 580s (read our review) were on fire sale. It’s as if VRAM has stopped advancing, and it is actively hurting gaming today. You’ll definitely notice this within a few years, but it’s also likely that it’ll affect the gaming experience for many people today.

If AMD wants to sell an 8GB “esports card” or whatever the company wants to call it, it can do that -- but do it for the true low-end. AMD is just right back to playing NVIDIA minus $50. AMD has to invest in its brand, like Intel is doing with Arc, where the company is willing to take less money to establish market share to try and eventually grow longterm. AMD could also shut the f*ck up on Twitter, specifically getting the executives off the social media platform. It’s not helping. The investments are supposed to go into products and getting market share, not into one-upping NVIDIA’s f*ckups. Giving out review samples doesn’t undo weaponization of incompetence from AMD marketing.

GPU Test Bench

RX 9060 XT Thermals

We’ll run a thermal test to evaluate how the Sapphire Pulse model does. These lower-end cards sometimes get shafted with poorly applied thermal solutions or just mounting issues.

In a 100% rendering workload, the 9060 XT GPU temperature sensor climbed to about 56 degrees Celsius steady state. That’s completely acceptable. The hot spot temperature was 79 degrees, meaning that there’s about a 23-degree delta between the two. The highest we’ve seen for GPUs has been in the 40s for a poorly installed GPU cooler, which is why we pay attention to the delta to begin with. In this instance, 23 is higher than we’d like to see, but both numbers are well within tolerances. We can thank the relatively low power limit of this particular GPU for that more than the cooler.

Memory temperatures ran warm at 85 degrees Celsius in an ambient of 21C. We’d like to see this cooler as it leaves little room for aging and ramping internal case ambient temperatures, but there’s still enough room here that it’s OK overall. We evaluate this closer in our tear-down video.

One thing we noticed was the bump right at the start: The fans took a little bit to kick in and didn’t hit until the GPU bounced off of what we assume is the VBIOS temperature target of 56 degrees for the GPU. After this, they aggressively cooled down the card, then slowly ramped back to that temperature target.

Frequency Target

Next, we checked the frequency log to ensure the card is hitting the advertised clock speed. In the same test where the GPU leveled-out at 56 degrees Celsius, the core frequency held at about 3130 MHz or so. The advertised boost frequency is “up to” 3130 MHz according to AMD, so it’s hitting that marker.

RX 9060 XT Acoustics

We also ran a quick acoustic test in our hemi-anechoic chamber that we built. The RX 9060 XT Pulse ended up almost being too quiet to get a good read on, as it’s near our noise floor. The device measured at about 16 dBA under full load and tested at 1 meter distance in the chamber, shown by the green line in this plot with some 9070 (read our review) models for reference.

The 9060 XT Pulse ended up with a slight spike at 500 Hz, which we’ve seen in the other cards on this chart, and then otherwise a relatively gradual frequency falloff later on in the plot. Nothing jumped out at us here, which is a compliment for the Pulse. The low power draw of the card allows it to spin relatively slowly. It is likely that the other fans in your system would be louder than this GPU.

RX 9060 XT Game Benchmarks

Dragon’s Dogma 2 - 4K

Dragon’s Dogma 2 is up first. Tested at 4K, the RX 9060 XT outperforms the RTX 5060 GPU. We’re in unplayable territory, but this is still a useful tool to understand scaling across resolutions.

The RX 9060 XT leads the 5060 by 7% here, with both behind the former flagship 2080 Ti GPU. The jump from the 9060 XT to the 9070 is huge, at 79% improved to the next card up. AMD has a large performance gap between these models. In between, we can find the 7900 XT (read our revisit) and 7800 XT (read our review) also from AMD, with the 6950 XT around the levels of the 7900 GRE (read our review). NVIDIA competition additionally includes the 5060 Ti (read our review), which ran at 40 FPS AVG and outperformed the 9060 XT by 11%.

Dragon’s Dogma 2 - 1440p

At 1440p, the RX 9060 XT ran at a playable 65 FPS AVG with lows proportionately spaced. The lows for all the latest NVIDIA and AMD cards are indicative of consistent frametime pacing for this game.

The 65 FPS result positions the 9060 XT near the level of the 7700 XT and 3070 Ti, just below the 5060 Ti. The 5060 Ti has a 7% lead, with the 9060 XT leading the 5060 non-Ti by 14%. The 9070 is significantly better, at 106 FPS AVG to the 9060 XT’s 65 FPS AVG, so there is a large gap between these. Currently, that gap is filled by the 7800 XT, 7900 GRE, and 6950 XT of prior generations from AMD, or the 3080, 4070, and 5070 (read our review) from NVIDIA.

Dragon’s Dogma 2 - 1080p

At 1080p, the 9060 XT again sits around the 3070 Ti and 7700 XT levels of performance. That positions it ahead of the RTX 5060 by 13%, with the 5060 Ti ahead of the 9060 XT by 8%. The 9070 improves to 134 FPS AVG from the 86 FPS AVG of the 9060 XT, or 56%, with the 9070 XT (read our review) improving by 68.5% to 145 FPS AVG, just below the 5070 Ti and prior XTX flagship.

Generationally and by name only (which, again, has shifted in meaning for both AMD and NVIDIA), the 9060 XT improves on the 7600 (watch our review) by 42% and the 6600 XT by 48%. The 6600 XT launched at around $380, but was available later in its life commonly for $220 to $260.

FFXIV 4K

Final Fantasy 14: Dawntrail is up now, first at 4K.

By framerate alone, the RX 9060 XT is basically the exact same product as the RX 6700 XT here. The two are nearly identical in performance, though there is a spec difference, most notably with VRAM. The original MSRP for the 6700 XT was around $480, and later in its life it was often available for around $300 to $350. In other words, against only the launch price, AMD has come down about $130 to equal the performance of a card from 4 years ago. Listings for a used 6700 XT range from about $250 (sold) to about $400, with common options in the $280-$320 range.

The 4060 Ti is also about tied with the 9060 XT, with the 7700 XT slightly ahead. The 5060 leads the 9060 XT in this game and benchmark, part of our established fact that AMD just generally struggles in this benchmark right now as compared to NVIDIA. In fact, even Intel’s B580 (read our review) is outperforming the 9060 XT. In this test, the 9060 XT just doesn’t do that well overall. The same was true of the 9070 and 9070 XT in this testing.

FFXIV 1440p

At 1440p, the 9060 XT ran at 85 FPS AVG, which has it equivalent to the 6700 XT from 2021, slightly below the B580 (although with better 0.1% lows), and just barely ahead of the RTX 4060 Ti. This game is one of Intel’s more promising for performance gains against the incumbents and remains one of AMD’s weaker games.

The 5060 leads the 9060 XT by 2% for AVG FPS, with lows functionally the same. The RTX 5060 Ti at 104 FPS AVG leads the 9060 XT by 22%. As we’ve seen in the past, AMD’s competitive performance in FFXIV just isn’t as strong as in other titles.

Generationally, the 9060 XT leads the RX 6600 XT’s 63 FPS AVG by 36%, the 6600 by 62%, and NVIDIA’s once most common GPU, the GTX 1060 6GB, by 158%.

Starfield - 4K

Starfield is up next. Sometimes, people ask why we still test this game. Well, we figured it made sense to benchmark a game nobody plays with cards nobody wants to buy.

AMD’s RX 9060 XT ran at 39 FPS AVG, with lows proportional to the average. None of the NVIDIA or AMD cards on this chart have particularly strong or weak 1% and 0.1% lows. The RTX 5060 Ti 16GB at 41 FPS AVG outperforms the 9060 XT 16GB by 5%. 

Against the new RTX 5060 that we bought for $330, the RX 9060 XT leads by 18.8% in average framerate. Lows scale along with the average. We need to drop resolution for more cards to compare.

Starfield - 1440p

Dropping to 1440p, the RX 9060 XT ran at 62 FPS AVG and ended up between the 5060 Ti and 3070 Ti. The RTX 5060 Ti’s 65 FPS AVG is indistinguishably different from the 9060 XT from an experiential standpoint, but technically ahead by 6%.

Buying a higher-end card instead, like the 9070, would improve the performance in this title by 56%, from 62 FPS AVG to 96 FPS AVG. One of the cheapest RX 9070s in stock at the time of writing was $650, about $100 over MSRP, which means that a $350 9060 XT 16GB (if you could find it at that price) would achieve 64% of the performance for 54% of the price. We expect it’ll be more than $350.

Down the stack and against older generations, the 9060 XT distances itself from the 6700 XT this time (despite similar performance in other games), and improves upon the 6600 XT by 63%, the 6600 by 94%, and the 2060 by about the same. Intel’s B580 is similar to the RX 7600 in performance, but suffers in frametime consistency in this game. Its experience would be noticeably worse as a result.

Starfield - 1080p

Tested at 1080p, the 9060 XT ran at 77 FPS AVG and maintained consistent lows, just like the other AMD and NVIDIA cards at its flanks. The 5060 Ti runs just 4.7 FPS AVG higher framerate, or a 6% advantage. Against the 5060 non-Ti, the 9060 XT leads by 17%. 

AMD’s prior 7900 GRE, 6950 XT, and NVIDIA’s new 5070 all end up ahead of the 9060 XT in a similar clustered ranking. Beyond that, the 9070’s 118 FPS AVG is 52% ahead. By MSRP, which the 9070 still regularly isn’t, that’d be a 52% performance gain for a 57% price increase by MSRP, but we can’t be sure where the 9060 XT will land until launch.

Resident Evil 4 - 4K

Resident Evil 4 is up now. At 4K, the 9060 XT’s 56 FPS AVG has it about equal to the 5060 Ti. Neither has a frametime consistency advantage over the other and all of these metrics are within variance. They’re the same result. The 3070 Ti is just behind the 9060 XT, with the 7700 XT just ahead of the 5060 Ti and 9060 XT. AMD improves upon the RTX 5060’s 46 FPS AVG by 20%, which is the same improvement from the Intel A770 (read our revisit). In this game, the Arc GPUs do better for frametime consistency than they have elsewhere. The 9060 XT is also 22% improved over the Intel B580, although the B580 is commonly cheaper.

Generationally, noteworthy markers in this chart include the RTX 2060, RX 6600, RX 6600 XT, and RX 7600, all of which would be significantly improved upon with a 9060 XT or 5060 Ti-class GPU.

Resident Evil 4 - 1440p

At 1440p, the RX 9060 XT held a 106 FPS AVG, putting it between the 5060 Ti and 3070 Ti. The 5060 Ti is 6% ahead in average framerate. The 7700 XT and 4070 outperform both of these, with the 5070 notably ahead both in price and performance.

The 9070 again leaves a large gap between it and the 9060 XT, with a 62% improvement over the new card. Versus the 5060, the 9060 XT runs 17% ahead, about the same as its uplift over the 3070, 4060 Ti, and similar to the 6700 XT. Intel’s B580 does OK to hang on here, but isn’t keeping up with the newer 5060 and 9060 XT. It’s getting closer though, so we’ll see what Intel achieves in the next generation.

Versus older cards, users of the 2060 would see an improvement of 127% to the 9060 XT, 6600 owners would see the jump from 58 FPS to 106 FPS AVG, nearly doubling, and 6600 XT users would also see meaningful uplift.

Black Myth: Wukong - 4K

Black Myth: Wukong is up now, tested first at 4K just for scaling purposes. 

The 9060 XT ran at 25.6 FPS AVG here. That’s obviously not playable, but let’s look at scaling: Against the 5060, performance is about the same. The 5060 Ti is 22% ahead, so similar to what we’ve seen even in higher framerate scenarios. The 9070 improves on the 9060 XT by 61% here. Let’s move to a lower resolution.

Black Myth: Wukong - 1440p

At 1440p, the 9060 XT climbs to 49 FPS AVG. That has it at about the same level as the 7700 XT and 3070 Ti. The 5060 Ti is about 17% ahead of the 9060 XT in this test, with the 7800 XT similarly ahead. As for the new 5060, that provides functionally the same experience when not running into VRAM issues. The 3070 is around the same spot as the 5060.

The 9060 XT at least improves over prior generation parts in a meaningful way: Against the RX 6600 XT’s 29 FPS AVG, we saw a 66% improvement. Versus NVIDIA’s old and popular RTX 2060, the uplift is 94% from 25 FPS AVG.

Black Myth: Wukong - 1080p

1080p puts the 9060 XT into playable territory without any form of upscaling and with these quality settings. The 9060 XT ran at 71 FPS AVG with lows where we’d expect them. The 5060 Ti’s 81 FPS AVG has it 14% ahead of the 9060 XT here, with the 5060’s 66 FPS AVG giving up an 8% lead to the 9060 XT. This is a stronger title for NVIDIA even when rasterized.

Intel’s B580 does OK for frametime pacing this time, but ends up far down the stack and adjacent to the old RTX 3060 and prior generation RX 7600. It’s just not competitive here.

Dying Light 2 - 4K

Dying Light 2 is up now. 

4K isn’t worth much time: First of all, as we all know, the human eye can only see 30 FPS without running two of them in SLI, which enables 60 FPS visibility. One of our viewers with glasses in the last review pointed-out that his spectacles enable MFG 4X, multiplying the visible framerate by a further 2X on top of SLI eyeballs.

All this to say that, at 32 FPS AVG for the 9060 XT, clearly we’re past perfection anyway.

The 5060 Ti outperforms the 9060 XT by 20% here, similar to some of the other tests even in spite of the heavy workload. That’s why we include these tests, though. The 5060 is about the same. Let’s move on.

Dying Light 2 - 1440p

At 1440p, the 9060 XT ran at 63 FPS AVG with good lows -- but not any better than any of its competitors’. The 3070 is slightly ahead and, critically, the B580 is actually about the same. Intel does well in this one, outperforming the 5060 in average and lows, while being functionally equal to the 9060 XT’s metrics. That’s more exciting than NVIDIA or AMD here.

But as for NVIDIA: The 5060 and 9060 XT would feel about the same in play in this game, with the 5060 Ti improving on the 9060 XT by 17%. Users of the now venerable RTX 3080 10GB can feel pretty good about their purchase, because the card is still hanging in there and keeping pace with modern GPUs, not too distant from the RX 9070.

Speaking of, the 9070’s uplift over the 9060 XT is 68%, putting them in totally different price and performance classes.

Cyberpunk 2077: Phantom Liberty - 4K

Cyberpunk 2077: Phantom Liberty is up next.

Tested at 4K, the RX 9060 XT ran at about 30 FPS AVG, so although there’s only a 6% lead for the 5060 Ti, we’re too constrained in performance to have a full picture yet. 

The 5060 fell notably below the 9060 XT here, with the AMD device holding a 21% advantage. Let’s see if that sticks at 1440p.

Cyberpunk 2077: Phantom Liberty - 1440p

At 1440p, the 9060 XT ran at 64 FPS AVG and landed between the 3070 Ti and 5060 Ti. The lead over the 5060 non-Ti drops to 16.7% here, down from about 21% at the less playable 4K. As for the 5060 Ti, its lead actually didn’t change much: It’s at 7.5% ahead now, from around 6% previously.

The 9060 XT is playable with these settings. Moving to something like a 5070 or 9070 would obviously be a huge jump up and would give some more room for higher settings or just future games, or even mods in this game, but that’s also a big price jump.

Cyberpunk 2077: Phantom Liberty - 1080p

At 1080p, the 9060 XT pushed up to a 100 FPS AVG. The 5060 Ti’s lead is now 8.8%, so we’ve seen it slowly increase its advantage over the 9060 XT as resolution has decreased. That’s why we run the 4K tests even when performance is too limited to play. As for the 5060, the 9060 XT leads it by 12% now, so it has consistently dropped from the 4K lead of about 21%. 

RX 9060 XT Ray Tracing Benchmarks

Ray tracing benchmarks are next.

Ray Tracing - Black Myth: Wukong 4K

We’ll start with the worst case scenario for everyone except NVIDIA, which is Black Myth: Wukong.

First at 4K upscaled, the 9060 XT ends up at the bottom of the stack. AMD has at least improved its ray tracing performance over the 7000 series, shown with the 7900 XT equaling the 9060 XT, but it’s not good enough to contend with NVIDIA (even if ignoring the low framerate at this resolution). The 5060 runs a 52% higher average framerate than the 9060 XT here. Let’s move to something that might actually run.

Ray Tracing - Black Myth: Wukong 1440p

At 1440p upscaled and ray traced, the 9060 XT ran at 31 FPS AVG and landed just ahead of the 2080 Super. That gives the 5060 non-Ti and its 45 FPS AVG a lead of 44%, down from the 52% previously. We’ll see if that continues to the more playable next resolution. The 5060 Ti ran at 54 FPS AVG, leading the 9060 XT by 72%. Even the 9070 XT was technically below the 5060 Ti, showing just how this particular game remains undefeated for NVIDIA.

Ray Tracing - Black Myth: Wukong 1080p

At 1080p upscaled, the 9060 XT ran at 47 FPS AVG. As always, for us, it’s not about the absolute framerate but instead about the relative framerate. In a relative sense, the 5060 ends up 37% ahead of the 9060 XT. In an absolute sense, it’s also just at a more playable FPS. The interesting part is the shrinking of the gap at lower resolutions here: Possibly because of how overrun it was, AMD’s GPU went from yielding a 52% advantage to the 5060, to 44% at 1440p, to 37% at 1080p. At this rate, maybe they’ll be at parity at 144p or something.

The 5060 Ti ends up 60% ahead here, with the 9070 XT around the same level. AMD gets left behind in this particular game with ray tracing on.

Ray Tracing - Dragon’s Dogma 2 4K

Dragon’s Dogma 2 with RT is up next. At 4K and not upscaled, just at normal 4K, the 9060 XT ran at 32 FPS AVG, basically tying the 2080 Ti and just behind the 3070 Ti. The 5060 Ti leads by 11% here, with the 9060 XT leading the 5060 non-Ti by 6.4%. Better performers include the 5070 at 49 FPS AVG, or 53% improved, and the 9070, at 76% improved. That breaks rank from what we saw in most rasterization tests. Of course, you could keep scaling up and we have results for that as well.

Ray Tracing - Dragon’s Dogma 2 1440p

At 1440p with RT and no upscaling, the 9060 XT’s 56 FPS AVG puts it well into playable territory with these settings. That has the 7700 XT just barely ahead, but the same from an experiential standpoint, and the 3070 is behind. The lead over the 5060’s 50 FPS AVG is 12% for a somewhat meaningful advantage. The 5060 Ti has a similar lead of 10% over the 9060 XT.

Compared to prior generations, AMD has improved on its architecture significantly and boosted its RT performance meaningfully. This is something we already saw in the 9070 XT review, but it continues at the lower end: The 9060 XT is a big jump over cards like the 6600 XT and 6600 previously.

Ray Tracing - Dragon’s Dogma 2 1080p

At 1080p, the 9060 XT’s 78 FPS AVG put it between the 7700 XT and 5060 Ti. The 5060 Ti’s lead drops from the 10% we saw at 1440p to just 5.5% at 1080p, with the 9060 XT increasing the gap against the 5060, now at 17% from 12% before. AMD is scaling better in this test as resolution decreases, even with ray tracing enabled.

Ray Tracing - Dying Light 2 1440p

We’re moving on to Dying Light 2 with ray tracing at 1440p upscaled. 

The 9060 XT was at 51 FPS AVG here and didn’t have any issues to speak of for frametime consistency. That has it ahead of the 5060 by less than 2 FPS, which wouldn’t be noticeable in real play. The 5060 Ti’s lead is more meaningful, at 17.8% over the 9060 XT. Overall, AMD is mostly improving on its own past performances here: Nearly hitting 7800 XT levels with the 9060 XT so at least that’s an improvement in RT.

Ray Tracing - Dying Light 2 1080p

At 1080p, the 9060 XT’s 77 FPS AVG has it again at about the level of the 7800 XT, just ahead of the 7700 XT. The 5060 Ti is about 13% better in average framerate, with the 9070 48% better than the 9060 XT and similar to the 5070.

AMD’s 9060 XT is 7.8% ahead of the RTX 5060 non-Ti here. The B580 is down below that, though not far below, and has OK frametime pacing this time. The 3060 Ti remains a weirdly relevant comparison to the 5060, which sort of feels like a 3060 Ti Super or something.

Ray Tracing - Resident Evil 4 4K

Resident Evil 4 ray traced is next, first at 4K upscaled. The RX 9060 XT 16GB ran at 63 FPS AVG here, putting it relatively close to the 5060 Ti’s 67 FPS AVG. The 5060 Ti has a lead of just 7% here, which is good for the 9060 XT. The 5060 gives the 9060 XT a 14% advantage, so it’s positioned better in this test than some of the other RT benchmarks. Lows are also consistent and proportional to the average, similar to the flanking devices from NVIDIA and AMD. Intel actually also had an OK overall performance here, with a 50 FPS AVG that begins to threaten the RTX 5060, though not the 9060 XT. Intel is mostly contending with the fact that it’s not consistently competitive, despite having improved this significantly with Battlemage.

Ray Tracing - Resident Evil 4 1440p

At 1440p upscaled and with RT, the RX 9060 XT’s 102 FPS AVG puts it just below the 7700 XT. The 5060 Ti is now 10% ahead, up from its 7% lead at 4K. The 9060 XT’s 10 FPS lead over the RTX 5060 non-Ti also reduces its advantage to 11% from 14% previously.

Ray Tracing - C77 VRAM Issues (1080p, RT Medium)

Finally, we’ll show the VRAM issues on the 8GB RTX 5060. These will also apply to the RX 9060 XT 8GB model.

In this ray tracing benchmark of Cyberpunk 2077: Phantom Liberty with 1080p/RT Medium settings, the RTX 5060 ran at 51 FPS AVG -- which sounds OK -- but had awful lows at 21 FPS for 0.1%. What’s important is looking at its neighbors: The RX 7800 XT had nearly the same average, but its lows indicated far more consistent and less choppy frametime pacing with the 43 FPS 0.1% result. The 9060 XT held a 54 FPS AVG and had well-timed, consistent frames, resulting in good 1% and 0.1% lows that represent an overall consistent experience. 8GB cards struggle here today, already, as shown with the 5060 and RTX 3070 Ti.

Ray Tracing - C77 VRAM Issues (4K, RT Medium)

To illustrate the point, here’s 4K with RT Medium. It’s ridiculous, but that’s the point: Despite struggling hard and overall being stuttery with its 15 FPS AVG, the 9060 XT had, to its credit, excellently paced frametimes for its average at 13.7 FPS and 13.6 FPS. There’s not a ton of data to work with here at this framerate, but the point is that even this sh*t experience is infinitely more playable than what we observe on cards like the RTX 3070 Ti, 3070, and RTX 5060, all of which are so variable from exceeding VRAM that the numbers are meaningless. We really can’t tell them apart: They’re all just unplayable in more ways than the 9060 XT is here.

So this is an illustration of VRAM limitations in real-time.

RX 9060 XT Efficiency Benchmarks

Up next, we’ll look at efficiency benchmarks when capturing GPU power consumption via the PCIe cables and the PCIe slot combined. This is isolated power draw to just the GPU. We’re representing efficiency in the form of FPS/W, or frames per joule, with higher being more efficient. The product name also contains the power drawn during the specific test.

Efficiency: Starfield 1440p

In Starfield at 1440p, the RX 9060 XT ended up slightly more efficient than the 9070 XT, though obviously with a lower framerate. The 9060 XT pulled 169W in this test, resulting in a 0.36 FPS/W result. The 4060 has a much lower framerate that’d push it into lower graphics settings, but is more efficient and has lower overall power draw. The RTX 5060 was also behind the 9060 XT in framerate, but also ahead in efficiency as a result of reduced total power draw of 126W for its framerate. The 5060 Ti pulled 145W and had a higher framerate than the 9060 XT, allowing it to approach the top of the charts with the 4060 Ti. Both of these are the 16GB models.

Efficiency: F1 24 RT 4K

F1 24 with RT is next and at 4K. The RTX 5060 couldn’t run this test without major stuttering, so it’s not on the charts.

The 9060 XT had a 0.17 FPS/W result, with the 5070 tying it also at 0.17 FPS/W, but with a 246W power draw. The 5060 Ti pulled 176W here. The 9060 XT is about the same power consumption as the 5060 Ti and only slightly behind in framerate, as this is one of AMD’s stronger RT titles. As a result, the efficiency is almost the same.

Efficiency: F1 24 RT 1080p

At 1080p with RT, the 9060 XT continued to pull about 178W. Its efficiency is better than that of the 9070 XT here, which is beyond the peak point for efficiency for the V-F curve for the 9070 XT. The 9060 XT is lower in framerate, but its lower power draw allows it this efficiency lead over its larger alternative.

The 5060 Ti ends up slightly more efficient than the 9060 XT, with both pulling around the same power and pushing a framerate within single digits of each other. This is a good showing for AMD’s new card.

Efficiency: FFXIV 1440p

We’ll look at a worse one for AMD as well. Final Fantasy XIV has poor overall gaming performance for the 9060 XT, which means the 9060 XT ends up lower down in the chart. It’s at 0.51 FPS/W, which at least has it improved on the 9070 XT and 7800 XT. The card is also tied with Intel’s B580, but related to NVIDIA, it’s behind the RTX 4060 and significantly behind the 5060 Ti.

RX 9060 XT Conclusion

8GB cards can exist, but we don’t think they should at $300. If AMD wants to sell an OEM piece of sh*t edition, the company is welcome to do that, but we think those should be relegated to true low-end devices. 

Selling the lower VRAM alternative opens up the door to less informed users, who are likely not in our audience, and getting them tricked into buying something worse than they’re led to believe. It also artificially deflates the MSRP when people talk about pricing.

As far as performance versus the RTX 5060, it’s at least better in nearly all instances. It’s often around levels of the 3070 Ti. It tends to be between the 3070 and 3070 Ti in a lot of tests. 

The 16GB 5060 Ti is currently $490 without a combo and without open box. The 5060 Ti is up to 14% better in our 1080p tests, up to 17% better at 1440p, and up to 22% better in our 4K tests. In ray-tracing games like Black Myth, it really pulls away, but that depends on the game.

As for price commentary, we’re not going to delve into that yet until we can see where the prices land. We’ll then do a GPU pricing comparison once things settle down as usual.

We test the RTX 5060 in several gaming benchmarks across 3 resolutions, which include rasterized and ray-traced tests

The Highlights

• NVIDIA did not seed out RTX 5060 GPUs for review
• The 5060 only has 8GB of VRAM, which will impose limitations in some scenarios
• Prospective buyers should also look at AMD’s RX 9060 XT
• Original MSRP: $300
• Release Date: May 19, 2025

Table of Contents

• AutoTOC

Intro

NVIDIA didn’t sample the RTX 5060, so we bought one while we were in Taiwan. Fortunately for our ability to get one, even a week after launch, they were in abundant supply and the stores told us that very few people were buying them.

Editor's note: This was originally published on June 3, 2025 as a video. This content has been adapted to written format for this article and is unchanged from the original publication.

Credits

Test Lead, Host, Writing

Steve Burke

Testing

Patrick Lathan

Camera

Mike Gaglione
Vitalii Makhnovets

Camera, Video Editing

Tim Phetdara

Writing, Web Editing

Jimmy Thang

Today, we’re reviewing the RTX 5060. We bought one in Taiwan and one from Newegg. The Newegg price was $330 and the Taiwan one was $360, so we’re ignoring MSRP and using the price we actually paid for purposes of this review. This is a post-launch review, which means we can use real price data rather than what largely turn out to be fake MSRPs for reviews. We’ll likely be buying all test samples of NVIDIA GPUs for some time, if not permanently, but that should make Jensen happy.

As quickly as possible: The 5060 is about the same as a 3070 (watch our review). The RTX 5060 Ti 16GB we tested is typically between 15% and 25% better than the 5060 across 1080p, 1440p, and 4K; the 5060 is typically slightly worse or slightly better than the 4060 Ti (watch our review); the 5060 is between 6% and 27% better than the 4060 (watch our review), commonly about 20%; the 5060 is also about 34% to 49% better than the RTX 3060 (watch our review) from 2021. But the VRAM is the big problem: 8GB becomes limiting at times, and will increasingly become a limiter.

There are a couple games where Intel’s B580 (read our review) performs at about the level of the new RTX 5060, but Intel still broadly has ground to gain. AMD’s RX 9060 XT will launch soon, so we’ll have a review of that shortly (Update: We have a review of that here).

In this review, we have data going back to the GTX 1060 6GB (watch our revisit) from 2016, which means the 5060 has added only 33% more memory in about a decade.

Let’s get into this quick review.

RTX 5060 Overview & Specs

The RTX 5060 remains readily available online.

It ranges from $300 to $360 commonly, but as of writing this, we’re seeing options at $300 and $320 again. These were out of stock when we bought ours.

The RTX 5060 has 8GB of GDDR7 memory on an anemic 128-bit bus, 3840 CUDA cores, theoretically 48 ROPs, but who knows if they’re actually there, and uses GB206 for the core. It’s also on a PCIe 5.0 x8 interface, so it’s populating half of the available lanes.

We purchased the Gigabyte Eagle OC model, which is really not that OC, but what makes it unique is that it uses a half-length slot.

This is about as cut down a model as NVIDIA could have possibly made while still calling it RTX. 

In our story about NVIDIA’s shrinkflation, we made this chart showing how NVIDIA’s modern RTX 4060 is similar to the GTX 750 in CUDA core percent relative to the larger configuration. In fact, its RTX 5070 was similar to an RTX 3050 (watch our review) by this metric. If we were to add the 5060 to this chart, it’d be at 15.6% of the full GB202 configuration and 17.6% of the 5090 (read our review) configuration.

This is hardly deserving of GeForce branding. 

GPU Test Bench

RTX 5060 Rasterized Game Benchmarks

Starfield Benchmarks

Starfield - 1080p

Starfield is up first.

At 1080p, the RTX 5060 ran at 66 FPS AVG and improved upon the 3060 Ti (watch our review) by an impressive, world-rending 3.1%. This is the improvement we came here for. This is why we buy GPUs. In just 5 short years and one dead AIB partner later, the EVGA 3060 Ti FTW3 has finally been unseated by a new 60-class card. Wow, that’s exciting.

Even more excitingly, the 5060 improves upon the 4060 by 4.1 FPS AVG. That’s such a big improvement that it’s almost noticeable. This 6.6% uplift isn’t something that you can get with just any technology: You have to actively try to produce this little of an improvement. If only we ran all of this through an arbitrary frame multiplier, it’d look more impressive.

Against the 6700 XT (watch our review), uplift is only 13%. Intel is still bested by NVIDIA: The B580 ran at 51 FPS AVG, yielding a 29% advantage to the 5060; however, the Arc GPUs have had issues with frametime pacing in this game that significantly hurt their viability beyond what the average FPS shows. You can see that in the 1% and .1% lows. Below that, the 3060 ran at 49 FPS AVG for a 34% uplift to the 5060, with the 2060 (watch our review) yielding a 73% improvement.

Starfield - 1440p

At 1440p, the 5060 ran at 52 FPS AVG and slightly trailed the RTX 3070. The lead over prior generations lines up like this: The 4060’s 47 FPS AVG is indistinguishable at a human level, though technically the 5060 is 9% better in average framerate. The same is true against the 6700 XT. Intel’s B580 is down at 41 FPS AVG and gives the 5060 an advantage of 27% (but again, with frametime pacing issues). The 3060’s 38 FPS AVG establishes a 37% improvement to the 5060 when spanning multiple generations, with the B570 (read our review) and A770 (read our revisit) both in a similar spot for the AVG FPS. Upgrading from an RTX 2060 in this exact test would produce a 72% improvement in average framerate.

But, unfortunately, for the 5060, the 4060 Ti and 3070 are both superior, and the 3060 Ti is functionally identical in performance to the new 5060. That’s just sad. The RTX 3060 Ti is almost too good by today’s standards, despite initial mixed reviews: It caused issues for the 4060 Ti at launch, and now it’s causing issues for the 5060.

Starfield - 4K

4K has fewer results because none of it really matters anyway.

The RTX 5060 ran at 33 FPS AVG in this test. What we care about is the relative comparison: The 5060 is beaten by the 3070 Ti (Watch our review) by 17%. The 5060 Ti is just past that marker at 41 FPS average, leading the 5060 by 25%. The 5070 (read our 5070 review) is next in the hierarchy, up at a 65% improvement. The RX 9070 (read our review) runs at about the level of the 4070 Ti (watch our review), but we’ll have to see where the 9060 XT falls soon.

Dragon's Dogma Benchmarks

Dragon’s Dogma 2 - 1440p

In Dragon’s Dogma 2 at 1440p, the RTX 5060 ran at 57 FPS AVG with lows at 45 to 47. Proportionally, these are well-spaced lows indicating good frametime pacing. The 5060’s average equates to the 4060 Ti 16GB’s from last generation, but with half the VRAM. The RTX 5060 Ti (16GB) outperforms the 5060 by 22% here. The 9060 XT would be interesting if we were allowed to show it already, but we might not have tested it yet, and if we had, we couldn’t show it.

We can show the 9070 though, and that’s at 106 FPS AVG with its own consistent frametime pacing. The lead over the 5060 is 86%. Pricing is also significantly higher, though.

The two-generation-old 6700 XT roughly equates the RTX 5060, as does the RTX 3070 also from a few generations back. We’re seeing sold eBay listings for around $270 to $320 for the RTX 3070, putting it in the same price class as a new RTX 5060.

The 3060 Ti ran at 53 FPS AVG, meaning it achieved 93% of the performance of the RTX 5060. The 3060 Ti was originally around $400 MSRP and today can be found in sold eBay listings for $150 to $260, with several selling around $230.

Intel’s Arc B580 ran at 44 FPS AVG in this test, allowing the RTX 5060 a lead of 30%. MSRP is $250 and the cards were commonly in the $400s for a while, but we’re seeing regular listings at $300 now. At price parity, in this game specifically, that makes the 5060 a better buy than the B580, but Intel is at least entering the conversation and it is competitive in some of the other games.

Against the last generation 4060, the 5060 leads by 27% at 57 FPS to 45 FPS AVG. Against the 3060 from 2021, the improvement is 45%. That’s not great for 4 years. Versus the 2060 from 2019, uplift is 87%.

AMD’s 9060 XT is what we need to really study in a few days.

Dragon’s Dogma 2 - 1080p

At 1080p, the RTX 5060 ran at 76 FPS AVG and held lows at 60 to 62 FPS. The 3070 again roughly equates the 5060, with the 3060 Ti only marginally behind at 68 FPS AVG. The 6700 XT ties the 3060 Ti, with the 5060 about 12% better than both. Generationally, the 5060 improves on the 4060 by 23%, the 3060 by 48%, the 2060 by 91%, and the 1060 by 237% -- although the GTX 1060 had a variant at 6GB, so somehow we’ve only improved on that by 33%.

Above the 5060, we have the 4060 Ti at a couple FPS ahead, which is just embarrassing for the 5060. 

The 5060 Ti leads the 5060 by 22%, with the 5070 leading by 65% and 9070 by 75%. These latter two are significantly more expensive, of course, so that makes sense.

Final Fantasy XIV Benchmarks

FFXIV 4K

Final Fantasy 14: Dawntrail is up now, first at 4K. This particular set of tests is a weaker showing for the AMD RX 90 series, as we’ve shown in the 9070 XT review and 9070 review. NVIDIA is mostly competing against itself.

The RTX 5060 ran at 47 FPS AVG when at 4K, landing it below the Intel B580. That’s great news for Intel, considering this is only their second generation. Lows are the same between them, meaning frametime pacing is the same -- another victory for Intel, considering how new it is here.

The RTX 5060 Ti leads the 5060 by a paltry 16%, with the 4070 28% ahead and the 9070 34% ahead. Again, the 9070 series has not done well in Final Fantasy 14 in this testing and that shows here. The other direction, the 5060’s 47 FPS AVG leads the 4060 Ti by just 5 FPS, with the 6700 XT just behind that. For reference, sold 6700 XT 12GB listings have the card between $200 and $320. The A770 is just below that, giving the 5060 just a 16% lead. Generationally, the 5060 leads the 4060 by 28%, the 3060 by 41%, the 2060 by 64%, and the 1060 by 191%.

FFXIV 1440p

1440p has the 5060 again tied with Intel’s B580, again in an embarrassing showing for NVIDIA. If Intel keeps at it, the company may have a real chance at disruption -- the hard part for Intel will be maintaining competitive prices while often using larger dies to be competitive.

The 3060 Ti leads the 5060 in another embarrassment for NVIDIA, at 3% ahead of the 5060. The 3070 and 7700 XT are next in line, followed by the 5060 Ti 16GB at 19% ahead. The gap over the 5060 increases at 1440p versus 4K based on these numbers.

Generationally, the 5060 uplift over the 4060 Ti is just 4%, or 21% over the 4060, 39% over the 3060, 56% over the 2060, and 163% over the 1060. All of these numbers are down versus the improvement at 4K.

Resident Evil 4 Benchmarks

Resident Evil 4 - 4K

Resident Evil 4 at 4K has the RTX 5060 at 46 FPS AVG, again basically tying the RTX 3070 -- but there’s something worse this time: The A770 also ties the 5060, with the B580 just behind. The RTX 5060 Ti leads the 5060 by 22% here. We’ll have to see if the 9060 XT falls anywhere near this area. For now, the 9070 is 96% ahead at a much higher base price.

Against older parts, the 5060’s 46 FPS AVG has it basically tied with the 4060 Ti and only 7% ahead of the 3060 Ti. The B570 sits down here too, with the 4060 down at 35 FPS AVG to give the 5060 a lead of 34%. Improvement on the 3060 is 55%, with the uplift on the 2060 at 92%.

Resident Evil 4 - 1440p

At 1440p, the RTX 5060 ran at 91 FPS AVG and kept frametime pacing consistent. Not in a remarkable way, but a predictable one: Most of the GPUs tested in this game fare well for frametime pacing, and the 5060 is among them. The 4060 Ti had a result functionally equal to the 5060, with the 6700 XT slightly ahead of both and in all 3 metrics. The 3070 Ti is next, with the 5060 Ti at 112 FPS and ahead of the 5060 by 24%.

Below the 5060, the A770 and B580 both encroach on its rank at about 85 FPS AVG. Unfortunately for Intel though, frametimes are inconsistent on these cards, as shown in the low variability for the B580 as compared to the A770 and 5060. Arc still has ground to gain here.

Versus lower down GPUs, the 5060 improves upon the 3060 Ti FTW3 by 10%, the B570 by 24%, 6600 XT by 28%, 4060 by 31%, and 2060 by 94%.

Black Myth: Wukong Benchmarks

Black Myth: Wukong - 4K

Black Myth: Wukong is next, which is a title that does particularly well with NVIDIA devices (and disproportionately so in ray tracing).

Predictably, 4K is unplayable on the RTX 5060. The 5060 ends up at 25 FPS AVG. From a relative standpoint, the 5060 Ti has a 23% higher average framerate, which is consistent with other tests. The 5070 is 59% better here and approaching playability, but isn’t there yet. The 9070 is around the same level as the 5070, including in 1% and 0.1% lows.

Let’s move on to something more realistic.

Black Myth: Wukong - 1440p

At 1440p, the RTX 5060 runs at 46 FPS AVG with frametime pacing consistent, not different from most other cards in this test (with the exception of Intel). The 5060 is about tied with the RTX 3070 and the RX 7700 XT. Realistically, the 4060 Ti 16GB is also at about the same performance as the 5060 here, but with way more VRAM. The 3060 Ti’s 42 FPS AVG means that the 5060 has improved upon it by just 9%. Generationally, the 5060 is 25% ahead of the RTX 4060, 47% ahead of the 3060, and 84% ahead of the RTX 2060.

Black Myth: Wukong - 1080p

At 1080p, the RTX 5060 held a playable 66 FPS AVG with lows representing consistent frametimes. The 7700 XT, 4060 Ti 16GB, and 3070 are all indistinguishable in experience for the player. The 3070 Ti isn’t much different. 

The 5060 ends up better than the 3060 Ti by only 13.5%. It improves on the 4060 by 21% and Intel’s B580 by 44%. 

Dying Light 2 Benchmarks

Dying Light 2 - 4K

Dying Light 2 rasterized is up next. At 4K, the RTX 5060 ran at just 32 FPS AVG and tied the RTX 2080 Ti from 2018. The RTX 5060 Ti leads the 5060 by 21%. We need to move down in resolution to see other cards.

Dying Light 2 - 1440p

1440p adds a lot of devices to the charts. The RTX 5060 is now at 62 FPS AVG, performing about the same as the RX 6700 XT and, again disappointingly for NVIDIA, the RTX 3060 Ti. This was also one of the games where the 3060 Ti embarrasses the 4060 Ti, so at least we see that trend continuing for one last generation. Even Intel’s B580 outperforms the RTX 5060, this time in both the average framerate and the lows. 

The 5060 Ti outperforms the 5060 by 20% here. Against the 4060 Ti, the 5060 is roughly equivalent (again), with the uplift over the 4060 at 26%, followed by the 3060 at 52% improved to the 5060.

Cyberpunk 2077: Phantom Liberty Benchmarks

Cyberpunk 2077: Phantom Liberty - 4K

Cyberpunk 2077: Phantom Liberty is up now.

At 4K and rasterized, the 5060 ran at 24.4 FPS AVG. This is only useful for a relative comparison. That’s the same as the RTX 3070 and slightly behind the 3070 Ti. The 5060 Ti is 28% ahead in average framerate here, with the 7800 XT ahead of that.

Cyberpunk 2077: Phantom Liberty - 1440p

At 1440p, the 5060 hits 55 FPS AVG in this test. That has it right between the 3070 Ti and 3070 and experientially the same. The 5060 Ti is better in average framerate by 26%.

Cyberpunk 2077: Phantom Liberty - 1080p

At 1080p, the RTX 5060 moves up to 89 FPS AVG and remains sandwiched by the 3070 and 3070 Ti. We haven’t re-run the 4060 through this test lately, but the 9060 XT will definitely be an interesting data point to see.

RTX 5060 Ray Tracing Benchmarks

Moving on to ray tracing testing now.

Black Myth: Wukong Ray Tracing Benchmarks

Ray Tracing - Black Myth: Wukong 4K

In Black Myth at 4K with upscaling, the RTX 5060 ran at 24 FPS AVG and roughly tied the 9070 -- which performs awfully in this specific game with RT -- and the 3070 Ti. This is unplayable, and limited VRAM will get the best of it as play time extends as well, so let’s move on.

Ray Tracing - Black Myth: Wukong 1440p

At 1440p with upscaling, the RTX 5060 ran at 45 FPS AVG and again roughly tied the 9070, which isn’t good for the 9070. This game continues to be an impossible challenge for AMD. The 5060 Ti leads the 5060 by 20% here, similar to what we saw in rasterized scenarios.

Ray Tracing - Black Myth: Wukong 1080p

At 1080p with upscaling, the 5060 now runs at 64 FPS AVG and is about equal to the 4060 Ti 16GB and 3080. The 4060 also appears on this chart at 52 FPS AVG, with the 5060 ahead by 23%. The 3060 Ti’s 48 FPS AVG means the 5060 improves on the Ampere card by 34%, or 77% ahead of the RTX 3060’s 36 FPS AVG. The B580 isn’t particularly competitive here, either, down at 34 FPS AVG and around the RTX 2070.

Dragon's Dogma Ray Tracing Benchmarks

Ray Tracing - Dragon’s Dogma 2 4K

Dragon’s Dogma 2 with ray tracing and tested at 4K has the RTX 5060 at just 30 FPS AVG, which we all know humans definitely can’t see faster than, and so this card is basically perfect. Just like Jensen said.

The 5060 Ti leads the 5060 by 18% here, but let’s move to 1440p for more results.

Ray Tracing - Dragon’s Dogma 2 1440p

At 1440p with ray tracing, the RTX 5060’s 50 FPS AVG tied it with the 4060 Ti 16GB and 3070. The 7700 XT seems like a particularly useful landmark for future benchmarks and reviews.

As for generational improvement, the 5060 posts a 27% improvement over the 4060, which was previously bested by the 3060 Ti, and a 47% improvement over the 3060 non-Ti. 

Ray Tracing - Dragon’s Dogma 2 1080p

1080p brings the 5060 up to 67 FPS AVG when tested with ray tracing. This has its performance experientially the same as the results for the 4060 Ti and RTX 3070, with the 3060 Ti not that different in experience at 60 FPS AVG. The 6700 XT is in the same boat as the 3060 Ti. Over last gen, the 5060 improves by 23% by name only. That’s ignoring NVIDIA’s name games it plays, where a 5060 is basically a modern 30 or 50-class GPU at best.

Dying Light Ray Tracing Benchmarks

Ray Tracing - Dying Light 2 1440p

Dying Light 2 with ray tracing and at 1440p upscaled is next. In this one, the RTX 5060 ran at 49 FPS AVG, with the 5060 Ti ahead by 22%. AMD does comparatively better in this one versus its Black Myth performance earlier, but we need to go down to 1080p for more cards.

Ray Tracing - Dying Light 2 1080p

At 1080p upscaled, the 5060’s 72 FPS AVG has it about the same as the 4060 Ti 16GB, but the 4060 Ti had superior 0.1% lows in this set of test passes. The 5060 is about 9-10% better than the B580 and 3060 Ti, so Intel is doing better here than in some other benchmarks. Its lows are also more consistent than some of its previous performances.

Against the 4060, the 5060 is 25% improved, or 56% improved over the 3060. AMD’s direct competitor will be on our chart within days, so we’ll check back for that. For now, the 7800 XT would be a reasonable comparison and outperforms the 5060 by 13%.

Resident Evil 4 Ray Tracing Benchmarks

Ray Tracing - Resident Evil 4 4K

Resident Evil 4 at 4K is up now, this time ray traced and with upscaling. The RTX 5060 ran at 55 FPS AVG, again roughly tying it with a 3070 and 6700 XT. The 5060 Ti is again around 22% better than the 5060 in average FPS here, with lows improved proportionally.

Intel’s B580 remains more competitive than NVIDIA probably wants it to be, but it still has some ground to gain to directly compete with NVIDIA when assuming price parity. It’s getting there, though. AMD’s direct competitor will be on the charts soon enough.

Ray Tracing - Resident Evil 4 1440p

At 1440p with upscaling and RT, the RTX 5060’s 92 FPS AVG has it about equal to the 4060 Ti 16GB and slightly ahead of the RTX 3070 and RX 6700 XT. The 5060 Ti leads by 22% again. Intel’s B580 is still back far enough that NVIDIA has some breathing room as long as Intel doesn’t make a big generational gain with Celestial, and also assuming its price remains competitive.

Against older hardware, the 5060 improves upon the 3060 Ti’s 81 FPS AVG by a little over 10 FPS, with the improvement on the 4060 at about 30%.

RTX 5060 Conclusion

Should you buy the RTX 5060 right now? No. You should also look at the 9060 XT, which offers 16GB and 8GB variants as well. Looking at the 8GB cards from both AMD and NVIDIA, we’re going to have the same complaints, which include running into limitations in at least some scenarios. At this point, 8GB cards at this class are getting ridiculous.

We believe that NVIDIA is damaging the PC gaming hobby and PC building. The company is doing a lot of shady sh*t right now and is acting in what we believe is an anti-consumer way.

We go over AMD’s Computex 2025 announcements which include the company’s new RX 9060 XT GPUs, Threadripper CPUs, AI Pro workstation GPU, and more

The Highlights

• AMD’s RX 9060 XT will have 8 and 16GB models
• AMD announced new Threadripper CPUs that include the 9980X, 9970X, 9960X along with PRO 9000 WX-series CPUs
• AMD also revealed the 9995WX, its new AI Pro workstation GPU, which will come with 128 AI accelerators, 32GB of GDDR6 memory, up to 1531 TOPS, and a 300W TDP

Table of Contents

• AutoTOC

Intro

AMD just announced its RX 9060 XT GPUs (coming in two memory configurations). We already knew about these but the company just formally announced them. AMD also revealed its “Radeon AI Pro R9700” workstation GPU, and the company’s latest Threadripper 9000 series and Threadripper PRO 9000 WX-series of CPUs. Unlike NVIDIA, AMD actually wants people to know about its products rather than about anti-consumer, anti-free-press actions, so AMD not only announced the products and gave information on them, but will also be sending out review samples well in advance and without conditions - which normally isn’t worth mentioning, but is worth pointing out because of the recent NVIDIA issues.

Editor's note: This was originally published on May 20, 2025 as a video. This content has been adapted to written format for this article and is unchanged from the original publication.

Credits

Host

Steve Burke

Camera, Video Editing

Mike Gaglione
Vitalii Makhnovets

Writing

Tannen Williams

Web Editing

Jimmy Thang

The 9060 XT 16GB will be $350 with the 8GB model at $300. They will release on June 5th. The GPU die is the same for both models, but new from the 9070 (read our review) and 9070 XT (read our review) (which also shared a GPU die). The new die is Navi 44 for the 9060 XTs and sized at 199mm^2, down from 357mm^2 on the 9070-class cards.

The company didn’t provide as many first-party testing results as they typically have in the past. That’s not necessarily a bad thing, because all of those results have to be taken with a grain of salt anyway, but we’ll mainly just be sticking to the specs today. We plan to review these cards once they launch. Our understanding is that, unlike the 5060, AMD plans to continue sampling GPUs as usual.

Just a heads-up: The information in this article is from a pre-briefing, so this is based on conversations with AMD and not the live presentation itself.

AMD Radeon RX 9060 XT

The 9060 XT will come in 16GB and 8GB versions. 

As for the features that will be shared between the two: Each will have 32 compute units, 32 hardware RT accelerators, 64 of what AMD calls its AI accelerators, and a 3.13 GHz boost clock. Both models will run PCIe gen 5.0 x16 slots, DisplayPort 2.1a, and HDMI 2.1b. AMD also lists a range of 150W-182W for board power, which explains the single PCIe 8-pin connector pictured in the rendering of the GPU. In speaking with AMD, the lower end of the range is for 8GB models.

For reference, AMD’s 9070 XT has 64 compute units, 64 RT accelerators, and 128 AI accelerators, or double the amount of the 9060 XT’s CUs and accelerators. The 16GB 9060 XT matches the memory capacity of both the 9070 and 9070 XT, but with a weaker core. 

These 9060 XTs will be direct competitors to NVIDIA’s 5060 Ti cards, even mirroring the same VRAM configurations. Despite identical memory sizes, NVIDIA’s 50 series cards utilize a newer GDDR7 memory compared to AMD’s GDDR6. As for how much that matters, that depends on the architecture and how much it’s going to rely on the memory bandwidth and that extra speed. We’ll look at it in testing and see how it performs in the real world. 

Additional differences include the 9060 XT’s use of PCIe 5.0 x16 as opposed to the 5060 Ti’s (read our review) PCIe 5.0 x8 interface. In benchmarking at x8 vs x16 on gen 5, it’s not going to matter. The place where it might matter is socketing it into an older board where cutting the lane count in half is going to become a restriction in some configurations.

AMD’s first-party benchmarks compared it against the 8GB RTX 5060 Ti, which we think is fair since it’s a price-parity comparison. We’ll do our own benchmarking pretty soon in our review. 

AMD Ryzen Threadripper 9000 and PRO 9000 WX-Series CPUs

AMD also announced its newest Threadripper 9000 Zen 5 CPUs, including the 9980X, 9970X, and 9960X, codenamed “Shimada Peak.” These have been upgraded with increased memory support and enhanced AVX-512 for more demanding tasks.

We’ll start with the standard AMD Ryzen Threadripper 9000 CPUs. This series includes the 9980X, 9970X, and 9960X.

The 9980X is a 64C/128T CPU at 3.2 GHz base clock and with 256 MB of L3 cache. The 9970X has 32 cores, 64 threads, a 4.0 GHz base clock, and 128 MB of L3 cache. And finally, the 9960X will come with 24 cores, 48 threads, a 4.2 GHz base clock, and also 128 MB of L3 cache. All of these CPUs will also feature an up to 5.4 GHz max boost clock, PCIe 5.0, the same sTR5 socket, and a 350W TDP.

For AMD’s Ryzen Threadripper PRO 9000 WX-Series: The company announced six new CPUs, which include the 9945WX, 9955WX, 9965WX, 9975WX, 9985WX, and the flagship 9995WX. Starting with the 9945WX and working our way up, these chips will come with core counts of 12, 16, 24, 32, 64, and finally 96 cores for the 9995WX. This mirrors the existing and prior 7000 series CPU configurations just now on Zen 5. 

Both the PRO and non-PRO Threadripper CPUs seem to resemble the same basic specs as the 7000 series of Threadripper processors. In these spec sheets, with a higher max boost frequency for the 9000 series CPUs, but a lot of the rest is familiar. 

One notable difference between the PRO WX and the non-PRO series of Threadrippers is that the workstation series offers “AMD PRO technologies,” which AMD describes as, “a robust suite of enterprise-grade features including multilayered security, advanced remote manageability, and long-term platform stability.” Additionally, at least in the past, the PRO WX-series cards supported the WRX90 chipset in addition to the TRX50 chipset.

AMD hasn’t announced any prices at this time, but the press-brief lists availability for July 2025, so we should be expecting to see these soon.

AMD Radeon AI Pro R9700

Finally, AMD introduced its latest AI Pro workstation GPU. Intel also just announced its new Pro GPUs this past week and we already have a tear-down up of the B60.

For specs, this RDNA 4 card will come with 128 AI accelerators, 32GB of GDDR6 memory, up to 1531 TOPS claimed, and a 300W TDP.

Compared to one of its predecessor, the Radeon Pro W7700, the new R9700 increases TFLOPS (FP16) from 56.54 to 96, increases AI accelerators from 96 to 128, upgrades to PCIe Gen 5 from Gen 4, and doubles the memory size from 16 to 32GB. Unfortunately, AMD’s press-brief didn’t include any CU, stream processors, or memory bandwidth info for the R9700, so we’ll have to wait to see those exact specs.

Due to the R9700’s noticeable configuration improvements over its predecessor, the new GPU ends up being slightly more comparable to the Radeon Pro W7800 which has 140 AI accelerators, 32GB of GDDR6 memory, and 90.5 (FP16) TFLOPS. 

In its press-brief, AMD included a slide to illustrate how 32GB of VRAM gives users more options in their ability to load larger AI models by highlighting four models that would exceed 16GB of VRAM, but can be used with 32GB of VRAM instead. Additionally, due to the GPU’s ability to load models with larger parameters or that are less quantized, the GPU may also see an uplift in the accuracy of the model’s responses.

To expand upon that point, AMD includes another chart labeled “Large AI Models Performance” where it compares an RTX 5080 (read our review) to its AI Pro R9700. Once again, this chart demonstrates how 32GB offers access to run larger models that 16GB just can’t handle. These results are expected. We think a more meaningful comparison might’ve been using the RTX 5090 that also has 32GB of VRAM. This would represent a more like-for-like scenario but we don’t do a lot of ML testing so we’ll leave that for someone else. 

AMD also shows off the card’s “Multi-GPU PCIe 5 platform” that allows users to connect 4 AI PRO R9700s for some extremely demanding models that need up to 128GB of combined VRAM and theoretically 4x the computing power.

We didn’t receive a price for this card, but AMD lists an availability of July 2025. 

Conclusion

That’ll wrap it up for AMD’s announcements.

Unfortunately, it’s a bit difficult to get an idea for performance based on the specs alone, and even harder to get an idea for the value for something without a price.

Ideally, we’d be able to get our hands on some of these once they’re publicly available, which should be soon according to AMD’s press-brief.

We compare AMD's in-game frame generation, AFMF, and FSR against “native” rendering in many games

The Highlights

• Our tests were performed using an AMD RX 9070 XT and RX 7900 XTX
• With frame generation, you still get some really nasty images that aren’t representative of how the game was intended to look
• FSR 4 generally improves image quality and stability over prior versions

Table of Contents

• AutoTOC

Intro

Today, we're giving AMD the NVIDIA treatment: We're inspecting AMD's fake frames to compare them to real frames, but we aren't yet comparing NVIDIA's fake frames to AMD's fake frames, because that'll come later.

That means that this article will include frame-by-fake frame analysis of AMD's generated images versus native-only rendering.

Editor's note: This was originally published on April 8, 2025 as a video. This content has been adapted to written format for this article and is unchanged from the original publication.

Credits

Test Lead, Host, Writing

Steve Burke

Testing, Writing, QC

Patrick Lathan
Jeremy Clayton

Video Editing

Vitalii Makhnovets
Tim Phetdara
Andrew Coleman

Writing, Web Editing

Jimmy Thang

Sometimes, you can see ghost images, such as in Kratos' swing over this snowy background, where there's a blurring of the axe and arms as AFMF, or Advanced Fluid Motion Frames from AMD, interpolates in-between frames. 

We'll also talk about FSR 4 vs. prior FSR iterations and native: In some scenes, like the one above, image clarity and stability are greatly improved over prior FSR versions. 

The Ultramarine's armor and hanging cables both show significant improvement in the newer version versus the older.

In other scenes, like the one above showing an air assault, we can see heavy warping with FSR 3.1, but still modulation with FSR 4 for the flying units. 

The ground assault shows issues with shadows pulsing underneath the Tyrannids in both versions. 

Smeared trails behind NPCs and barrels are improved upon with FSR 4, but sometimes still present.

So, we'll be looking at AMD's Fidelity FX Super Resolution version 4 with the new RX 9070 XT (read our review) GPUs and comparing it against the prior version. This is an image quality specific test, and like we said in the NVIDIA coverage of DLSS and MFG, not all fake frames are created fake equal. This will look at that in part. What we're not doing here yet is comparing FSR 4 and AFMF to DLSS and MFG. That might be a later piece if there's interest, but we need to lay the groundwork for each technology independently first.

One important thing to remember with all of this, just like with NVIDIA’s that we looked at, is that we’re closely inspecting these images today for image quality. That means we’re pausing things and zooming in. In real play, it’s likely that some of these differences would go unnoticed at full speed and “zoomed out.” One other note is that YouTube/video compression makes things sometimes difficult to fully appreciate.

Let’s get into it.

FSR 4 Overview

AMD’s FidelityFX Super Resolution (FSR) upscaling has finally moved into the “AI” buzzword era with FSR 4’s machine-learned Convolutional Neural Network (CNN) model co-developed with Sony. Sony and AMD announced a collaboration effort back in December of 2024, dubbed “Project Amethyst.”

Sony strongly implied that it’s going to use a rebranded version of FSR 4 as its own “PSSR” in order to target 1080p native rendering on the PS5 Pro, but with the upscaler doing the work to output a good looking image at “4K.” 

On the PC side, FSR 4 will only run on the new Radeon RX 9000 series graphics cards for the time being, with no official word on back-porting to RX 7000. We’re unsure at this time whether it’s a technical limitation or a product segmentation move on AMD’s part.

The official support list for FSR 4 has 36 games at the time of writing, which is late March. That’s not a lot, but the number will hopefully grow as more games are updated. Several of the listed games are big Sony titles as well, indicating that the company is serious about utilizing the tech, but also shows the partnership between them.

The previous generation, FSR 3.1, is technologically distinct from its predecessor (FSR 3) by way of being implemented as a modular .dll file rather than being entirely baked-in to the game. This paves the way for future revisions of FSR to be more easily implemented by the game developers or just in general.

FSR 4 also uses a .dll file, and can be swapped-in officially in FSR 3.1 games via a driver-level override in AMD’s Adrenaline software in a very similar way to NVIDIA’s DLSS override. However, the games have to also be on AMD’s official whitelist to get the toggle to appear in the driver software. 

Unofficial tools like OptiScaler open the door for a lot more flexibility, but we haven’t tested them yet so we can’t make a recommendation, but there’s stuff like that out there.

AFMF 2.1 Overview

AMD also includes in-game frame generation, or “fake frames,” under its FSR umbrella, and we haven't seen any indication from AMD that its in-game framegen algorithm has changed since FSR 3.1. 

In AMD's words, "Advanced frame generation interpolation technology when used with AMD FidelityFX Super Resolution (FSR) 3 inserts 1 frame between existing ones."

However, AMD also has separate driver level frame generation known as Advanced Fluid Motion Frames (AFMF) that can be applied without in-game support. AFMF 2.1 is a new introduction alongside FSR 4.0. To use it, you need AMD Software: Adrenalin Edition 25.3.1 or newer, RX 6000 or newer, and a DX11, 12, or Vulkan game. RX 6000 only supports AFMF in exclusive fullscreen mode, while RX 7000 and newer support borderless windowed, and the AMD 9070 (read our review) reviewer guide stated that "in-game display setting should be set to borderless fullscreen mode." 

And that's a lot of rules, but keep in mind that Smooth Motion, NVIDIA's answer to AFMF, is exclusive to the RTX 50-series. 

NVIDIA has stated that "support for GeForce RTX 40 Series GPUs will be coming in a future update." 

We're focusing on AFMF 2.1 here, so this isn't a direct 1:1 equivalent to the piece we just ran on NVIDIA's in-game frame-gen, but we’ll also be looking at the frame generation performance for AMD. For this article, the performance we care about is image quality, and not the actual literal framerate performance. That will be a separate test along with potentially latency. 

This is an isolated test so that we can build foundational knowledge first, just like we did for NVIDIA. The direct comparison would be NVIDIA Smooth Motion versus AMD Fluid Motion Frames, but we’re focusing on just AMD today. That comparison may come later.

As for FSR testing, our FSR comparisons will focus on FSR 3.1 vs. FSR 4, with a couple references to native capture as an anchor. We captured everything at 4K resolution with FSR running at the Performance preset, meaning it’s upscaling from 1080p base render resolution. We disabled anti-aliasing, camera effects, and motion blur where possible to get the cleanest images we could.

The objectives today are purely image quality, not performance. We’ll be comparing frame-by-fake-frame image quality, FSR iteration quality, and looking at behavioral patterns in general.

Let’s get into the image quality comparisons.

Warhammer 40,000 Space Marine 2

First up is Warhammer 40K Space Marine 2. Like everything else we tested for this piece, FSR 4 support comes by way of the driver-level override. We used the High graphics preset, turned off camera shake, and set motion blur to off; however, we found the latter doesn’t actually work and motion blur persists regardless, but that’s a game thing.

Armory

We’ll start the comparisons with a scene in the Armory, or “Armouring Hall” in native Grimdark. Even before walking forward, the difference between FSR 3.1 and FSR 4 is stark. Static elements like the floor of the walkway that shift and shimmer heavily with FSR 3.1 are now stably locked-in. On top of that, the entire image is much clearer and more detailed. 

Examples of this include the floor, where we see improved image stability and clarity, the Ultramarine’s armor showing similar improvements in a side-by-side, the distant hanging cables, and the tech priests’ hoods. Distant candle flames that can’t even be made out with FSR 3.1 are visible with FSR 4, bringing them back into existence.

As we walk forward and the… we’ll call it a “kiosk,” drops down, the difference in clarity is so obvious it’s almost like looking at two different resolutions. No matter where you look there’s improvement. The fine details under the main monitor, the tech priest’s mask, and even just the general contrast and visual discernability of all the mechanical arms and tubes are all vastly improved. We double-checked our settings and confirmed that they were running as intended for a like-for-like comparison, so it really is just that much better in this example.

This is a very promising start for FSR 4.

Let’s compare FSR 4 to native 4K. Space Marine 2 forces either TAA or upscaling at all times (even at native), so we went with the default of TAA for the native capture. 

We also did this in our DLSS comparison, which showed how some scenarios, like Cyberpunk, can actually look worse than upscaling because of TAA.

The flickering and shimmering on the floor in Space Marine 2 we saw with FSR 3.1 is also present at native with TAA. That makes FSR 4 look even more impressive here – even at the performance preset – since it’s taming an undesirable behavior. We said this before, but we shouldn’t be seeing things that look worse with all defaults at native than with an upscaling technology. 

General detail before moving – like in the floor, walls, and priest hoods – is a toss-up between native with TAA and FSR 4 Performance mode. That’s simultaneously a critique of TAA and a praise of FSR 4. 

The green orb far in the distance looks very slightly better with native TAA. As the “kiosk” drops down, the level of detail between native with TAA and FSR 4 Performance is very close. However, we think FSR 4 actually comes out slightly ahead – most obviously in the round speaker-looking elements on top of a couple of the monitors.

Like we said in the DLSS piece, upscaling should never look better than “native,” and game developers shouldn’t be leaning on upscaling technologies in this way. The only reason it ever does look better is because of issues such as those with TAA, which is the default here.

Command

We’ll briefly look at another indoor scene in the campaign’s command bridge before moving on to a mission. Again, everything is sharper, clearer, and easier to discern when using FSR4. Gadriel’s face, Titus’ hair, and Chairon’s armor – particularly the chest decoration – stand out as night-and-day differences.

Looking at the bridge’s holographic map area, FSR 3.1 suffers from distracting flickering on some of the round grates (underneath the three skulls and on the right side of the main terminal). Switching to FSR 4 almost entirely eliminates this behavior. The thin lines within the hologram that shimmer and boil with FSR 3.1 come across as much cleaner with FSR 4. The other ship visible out the left windows doesn’t show much difference, however.

It could be that the largest areas of improvement with FSR 4 are in dimly-lit or low contrast scenes. We’ll test that with the next comparison.

“Decapitation” Mission - Scene 1

Loading into the “Decapitation” Operations mission gives us a brighter outdoor area. While the marine’s armor is again visibly better-looking with FSR 4, the effect on the rest of the scene is more subtle, more like lifting a haze rather than a transformation. You can clearly see this in the texture of the ground and on the shoulders of the statues.

Once in motion, FSR 3.1 starts to deteriorate, but FSR 4 keeps the image clean. Unfortunately, the bugged permanent motion blur makes it hard to draw a distinction between the two FSR revisions while the marine is rolling down the stairs. Elements such as the detail in the inlaid stone floor and the buildings to the right look better during motion with FSR 4. So far, it looks like most of the FSR 3-to-4 improvements are visible in motion.

“Decapitation” Mission - Scene 2

Strafing to the side shows off a huge improvement favoring FSR 4 over FSR 3.1. Every single element on screen looks better with FSR 4 – the textures on the sides of the stairs, the statues, the murals against the far wall, and the guardsmen. It again gives the impression of running at a higher resolution.

This would be a very interesting scene to compare the most up to date implementations of DLSS and FSR head-to-head in a future piece.

“Decapitation” Mission - Scene 3

Last for Space Marine 2 is an upscaling torture test by way of a slow pan of the Tyrranid assault. The flying creatures passing over the front of the massive building cause the windows to warp and blur heavily when using FSR 3.1. It still happens with FSR 4, but to a lesser extent. This is close to a worst case scenario for any upscaler, period.

In both the air and on the ground, FSR 3.1 causes the creatures to blur into a mass, sometimes phasing in and out of existence or blending together. FSR 4 isn’t totally immune here either, but again does a way better job than the prior version.

Since this scenario is so hard on upscalers, let’s compare to native with its forced TAA again. Watching the flying creatures in front of the windows shows us what it’s supposed to look like, free of the heavy warping seen with FSR 3.1, or even the warping-lite modulation seen with FSR 4.

Looking at the ground assault again shows just how well FSR 4 is handling this relative to FSR 3.1. It’s very close to the look of the native capture, but still suffers from an effect that makes the shadows underneath the Tyrranids appear to pulse and shift underneath them. At this level of fine detail and chaotic movement, even native with TAA struggles a little bit with grain and warp. Some areas are cleaner, but like we said earlier, this particular game has toss-up comparisons between them.

Monster Hunter Wilds

Monster Hunter Wilds is up next. We used the High graphics preset, but turned off anti-aliasing and camera effects like motion blur, vignette, and depth of field. We also reduced camera shake to its minimum levels. Cutscenes are entirely in-engine, which is useful, so we got a mix of those with some actual in-game capture.

Title Screen

Taking a quick look at the title screen animation shows little difference between FSR 3.1 and FSR 4. Textures and edges are slightly sharper with FSR 4. The stippled effect on fur as seen with FSR 3.1 is also almost entirely gone with FSR 4. There’s not much else to discuss here so we’ll move on.

Oasis

Jumping into the actual game at the point you choose your weapon by the small oasis shows an ugly pattern superimposed over the sand when using FSR 3.1 that goes away with FSR 4. It doesn’t make the game unplayable or anything, but it’s pretty distracting.

When talking to Alma, we can see that FSR 4 handles the fine strands of her hair a little better than FSR 3.1 does. It also adds sharpness to the weapons immediately to her left. While running towards the training barrel, we see that FSR 4 keeps the detail of the hunter’s clothes and equipment sharper. It’s not a huge difference, however.

Going to Camp

The long cutscene that takes you to the main camp shows much the same. FSR 4 has a slight advantage to image quality via sharpness and minor detail enhancement. Hair, fur, and feathers benefit the most by way of FSR 4 reducing the appearance of patterns or stippling superimposed over them.

One clear difference is in trails behind objects moving across a patterned background. For example, when the two Felynes drop the barrel, it leaves a smeared trail behind it on the wooden walkway, as does the NPC on the left side of the screen and the unfortunate Felyne as it falls backwards. We think FSR 3.1 handles this very poorly – FSR 4, while not perfect, definitely does better.

Considering what we saw in Space Marine 2, the less drastic differences between FSR revisions in Monster Hunter Wilds is surprising.

Marvel Rivals

The final game we'll analyze for FSR upscaling is Marvel Rivals. To keep things consistent, we used the training range to gather footage.

Rocket Jump

Static scenes don’t differ much between FSR 3.1 and FSR 4 in Rivals, possibly because of the game’s art style and strong default sharpening filter. We’ll need to look at movement to suss out the differences, and blasting forward with Rocket’s dash is a good place to begin.

Even before dashing, we can see a blob of pixels around Rocket’s head getting distorted in the FSR 3.1 recording, which isn’t nearly as pronounced with FSR 4. As soon as we blast forward, FSR 3.1 turns into an over-sharpened, grainy mess, as seen on Rocket’s gun, jetpack, tail, and the ground below. FSR 4 is able to cope with the sudden movement far more gracefully, and only gets bad around the finer points of detail like the spikes on Rocket’s knees and the hair/fur on his head.

Rocket Strafe

Next we strafed to the side while shooting. Before moving, we can see a repeating pattern overlaid on the left side of the ground, similar to what we saw in Monster Hunter Wilds, that’s not present with FSR 4.

Once we do start moving and shooting, FSR 4 retains more detail in the bullet trails than with FSR 3.1. After dropping off the first ledge and reloading, FSR 3.1 responds to all the motion by giving everything in the vicinity of Rocket’s model a kind of deep-fried-meme look, and ghosting on the trailing edge of his jetpack. There’s still a little bit of ghosting with FSR 4, but it’s greatly reduced – and the deep-fried look is gone.

Rocket Wallrun

Last, we tried a wallrun. FSR 3.1 handled it overall better than we anticipated, but still left horrible ghost images as Rocket climbs up across the purple banner. FSR 4 still has them slightly, but they’re not noticeable to us at full speed in real-time.

For the rest of the run, there’s not a huge difference between the two FSR revisions outside of a slight clarity advantage and reduced ghosting with FSR 4.

AFMF 2.1 Image Comparisons

We’re getting into the AFMF and frame generation comparisons now. We’ll play some side-by-side, frame-by-frame comparisons while setting this section up.

For comparison purposes, we selected games that support in-game frame generation to compare with driver-level changes. In the real world, you should almost always opt for in-game frame generation over AFMF at the driver if in-game is available. The most obvious downside of AFMF is that it has no awareness of UI elements, so menus, text, and icons may be distorted (although we saw NVIDIA struggle with this even with in-game frame generation).

All footage for AFMF comparisons was captured at 4K 120FPS. A 7900 XTX (watch our review) with Adrenalin 25.2.1 was used for AFMF 2.0 capture and a 9070 XT with Adrenalin 25.3.1 was used for everything else. AFMF Search Mode was set to High and Performance Mode was set to Quality, the recommended settings for our setup. 

AMD is more direct than NVIDIA about recommending framegen only in scenarios where the game can already run at at least 60FPS, and we stuck to that recommendation.

As with NVIDIA framegen, the vendor's capture utility (Radeon ReLive in this case) was the most practical way to capture, but because we can't manipulate the rate of frame generation (like we could for NVIDIA), we'll have to rely more on frame-by-frame comparisons. We constantly had issues with AMD's Record & Stream tab disappearing after reboots, so if AMD is reading this, please fix that (and yes, our IGP is disabled).

Warhammer 40,000 Space Marine 2

We used the same settings for Space Marine 2 that we did for the FSR comparisons, with Resolution Upscaling set to FSR and Render Resolution set to Native since it can't be explicitly disabled. We set Motion Blur Intensity to Off even though the setting doesn't work. Framegen worked with fullscreen enabled, so we left that setting alone.

Screenshake

We’ll start with the worst-case scenario.

We recorded an additional scene in Space Marine 2 to show the downsides of frame generation. These clips aren't tightly controlled or synchronized; we just loaded into the main hub and shook the mouse around violently, so we’ll freeze frame some parts to show the issues. Besides making it very obvious that motion blur was still on (despite being toggled off), this allows us to see multiple frames where AFMF is definitely applied on top of UI elements like the "Assemble" waypoint. Applying frame generation on top of UI elements can cause ghosting and duplication of UI elements, which just looks bad. We can compare this to the in-game frame generation option, which is also a garbled, muddy mess in this scenario, with distortions around the edges that almost look like eye floaters. It completely breaks in this scenario, but perfectly preserves the UI layer even in the worst frames. 

Rapid movements like this aren't out of the question for mouse-and-keyboard users. They’re especially common in certain types of games, like shooters in particular where a fast or twitchy response necessitates them. Compared to native, these types of rapid movement scenarios are a worse experience with frame generation.

Armory

Our armory scene is relatively slow, but we can still see the effects of frame generation on moving limbs versus native rendering. The telltale sign of generated frames here is the slight blurring of detail on the marine's armor in frame-by-frame playback, like the back of his right leg as he passes under the light, but this is only really visible when closely examining individual frames in a specific location. The mandatory motion blur and application of FSR also help disguise the generated frames, since both of these effects lead to the same kind of temporal smearing that generated frames are subject to.

The frame-by-frame playback of AFMF 2.0 versus 2.1 reveals greater differences, with a prominent secondary image surrounding the marine's right arm in advance of its movement in the next "real" frame. This only affects moving elements of the scene; there's no noticeable sign of frame generation in the relatively static background even with the older version.

“Decapitation” Mission - Scene 1

By rolling down the stairs, we can see a weak point of frame generation, although the inescapable motion blur makes it harder to detect. There are multiple frames where the fine details on the marine's armor are blurred, like the bottoms of his boots where lines are lost. Moving frame by frame with native and AFMF 2.1 capture side by side, there's a clear difference between the even and predictable motion blur and the irregular, faded outlines of generated frames. The marine's feet change position more than any other part of his armor during the dive and roll, which is why they're especially prone to blurring and transparency.

The generated frames are also identifiable by ghosting behind the bullets being fired in the background, with classic ghosting behavior causing an undesirable look. In our recent piece discussing NVIDIA's framegen, we noted similar behavior with raindrops that made the rain effects look softer and more transparent in motion.

Comparing AFMF 2.0 to 2.1 at similar framerates, 2.1 appears to have less ghosting than 2.0, which should help make the effect less noticeable during gameplay. In each generated frame of the AFMF 2.0 capture, there's a clear secondary image of the marine in advance of the next real frame, as we noted to a lesser degree in the armory scene. Blurring and loss of detail can be issues, but ghosting is a much more visible downside of framegen and one that can ruin the experience. 

A comparison to in-game frame generation at the same timestamp demonstrates that it's predictably much better at preserving detail, since it's given more information to work with directly from the game engine. 

The in-game generation causes a crisp outline in advance of the marine's movement, visible above around his gun and his arms, but the detail within his boots isn't distorted at all. This could still lead to the shimmering outlines that we saw with NVIDIA's framegen, but the individual frames are closer to reality than with AFMF.

“Decapitation” Mission - Scene 2

We'll start off this next scene with another native comparison to establish which elements are due to regular motion blur and which are due to frame generation. The AFMF 2.1 capture matches native with its rendered frames, but it's interspersed with generated frames where the marine's armor loses sharpness on the fast-moving legs. That's clearly the area we need to focus on, so we can move to a 2.0 versus 2.1 comparison with that knowledge.

Walking sideways in front of a light background makes the difference between 2.0 and 2.1 more obvious, with 2.1 showing some slight irregularities around the edges of the marine's legs where 2.0 shows massive secondary images. As usual, the resulting effect would be easier to notice in motion if there weren't already other blurring effects forced on us with this game.

God of War Ragnarök

God of War ran well above 60FPS on the 9070 XT even at Ultra, which made it harder to capture generated frames, so we kept our testing brief. The ultra preset was used with motion blur, film grain, and camera shake disabled. We expected FSR to be forced when in-game framegen was enabled, but the options menu didn't reflect that, so the scaling method was set to TAA and the quality to Native for all tests (scaling cannot be explicitly disabled).

Axe

As expected, the moments of rapid movement during Kratos' windup and swing are the most difficult for AFMF to deal with. Whereas the native capture clearly shows each frame of his arms and the axe as it moves, the AFMF 2.1 capture has individual frames with obvious attempts at interpolation. This is most noticeable directly on the models, since the trail of particles behind the axe is already an intentional smear.

By swinging Kratos' axe in front of the white snowy background, we can clearly see ghost images with the older AFMF 2.0 frame generation. Direct frame by frame comparison is more difficult here due to the lower frequency of generated frames, but it's clear that the generated frames with 2.0 more commonly have ghosting around the axe and even Kratos' model. 2.1 shows some artifacts as well, but it's usually in the form of unevenness in the outline of the axe and Kratos' arm as they swing forward, while the core of his model is better preserved.

Checking back against God of War's specific implementation of in-game frame generation, we see the same behavior as in Space Marine 2 where the details of Kratos and the axe are excellently preserved, but there's a distinct outline where framegen has done its version of content aware fill in the area that those objects will occupy in the next real frame. In comparison, the fullscreen AFMF 2.1 effect has more distortion around the usual axe and arm area.

Log

The movement of the axe in this small QTE was too rapid for either 2.0 or 2.1 to keep up with, with the large deltas between frames leading to a similar appearance from both AFMF versions. The head of the axe appears doubled in some generated frames, with additional distortion around the handle. It's difficult for any type of frame generation to compensate for large deltas between frames, which is part of why AMD recommends running at 60FPS before turning it on (with latency being the other part).

Marvel Rivals

Marvel Rivals is next.

Mantis 3

Starting with the jump pad, we can see at the beginning of this scene that both AFMF 2.0 and 2.1 show ghosting on the rings that rise up from the pad. Once in motion, both iterations have trouble keeping up with the rapid movement, but 2.1 does a moderately better job of preserving Mantis' outline. 

The distance between the secondary image and the core of Mantis' model with 2.0 creates a greater blurring effect, and although the generated frames with 2.1 only roughly maintain the shape of her hands and feet, the resulting silhouette is stronger. As she approaches the apex of the jump, the distortion occurs in the background instead, with 2.0 more commonly showing doubling of scenery elements like the railing immediately in front of her.

Marvel Rivals' in-game frame generation version behaved in the same way as God of War's and Space Marine 2's,  with crisp outlines around the next known position of Mantis' model, visible above the knees in these frames, coupled with some afterimages trailing behind the arms. The overall effect is still far cleaner than AFMF, especially since the artifacts only apply to Mantis' model, leaving the scene behind her almost completely clear of framegen problems.

Mantis 2

We don't recommend using framegen for games like Marvel Rivals where latency is a concern and rapid mouse movements are common. Framegen is kind of like v-sync in that at best it's a cosmetic upgrade maybe, and at worst it's a source of latency. 

In this scene, we can see distortion around Mantis' hands with AFMF 2.1 in advance of the motion they're about to make as she fires, but a rendered frame was buffered in order to generate that image, whereas without framegen, rendered frames are simply delivered when they're ready.

Notice also that with AFMF, Luna Snow's nametag is distorted on the first frame where the camera shakes, which is a further disadvantage of AFMF versus in-game frame generation. Using the in-game option, the nametag is preserved.

Cyberpunk 2077

In Cyberpunk, we used the High preset without ray tracing and set FSR 3 to Native AA, since FSR was required in order to enable in-game framegen.

Car

First is a car scene. Keeping an eye on the headlight beam as it passes over shrubs and rocks to the right of the road, we can see that the scenery appears to wiggle slightly with AFMF 2.1 as the generated frames don't place objects in precisely the correct location. With AFMF 2.0 in the same areas, there's a much simpler loss of detail as every generated frame is a blurry mess. The area on the shoulder that tracked slightly wrong with 2.1 is instead completely layered with a ghost image with 2.0; we prefer the newer version here, but native rendering without fake frames still looks cleanest.

Moving to a 2.1 versus native comparison, we can see that the wiggling behavior is definitely an artifact of AFMF and not something carried over from the original native render. The native capture has fewer frames, obviously, but the objects to the right side of the road track in a clear, straight line between frames as opposed to the slight side-to-side shifting with AFMF 2.1. Native should typically be crisper, even if less smooth, so this makes sense.

Benchmark

Using the game's built-in benchmark scene, we can again see that 2.1 is more prone to distortion in areas where 2.0 would blur and ghost, with the vertical pillar here turning into a wavy line in fake frames. 

As the camera passes the corner with pages from the Night City Journal, 2.1 does a much better job of preserving text and fine detail where 2.0 just ghosts. 

Moving to a later part of the scene, frame by frame playback reveals more distortion visible with 2.1 as the camera passes over the barbed wire fence: straight lines remain clear with 2.1, but they don't remain straight. As usual, 2.0's equivalent is a fully duplicated ghost image of the fence in each generated frame.

Again, we can check back against the native footage for confirmation that the behavior we're seeing comes from AFMF. The horizontal fence bars are straight and uninterrupted in each frame of the native capture, whereas they're frequently broken up and uneven in the AFMF 2.1 footage, although AFMF 2.1 does do a fairly good job of preserving the scene behind the fence without major framegen distortion.

Conclusion

In terms of image quality, FSR 4.0 is an improvement over FSR 3.1. And AFMF 2.1 is an improvement over 2.0. Both of these are good things because if those weren’t the case, then AMD is using numbers wrong. 

Whether or not it’s worth using these technologies at all is more situational, just like with DLSS. First of all, it’s going to be highly specific to the games and also the person playing them. The use case for these, take FSR for example, is definitely a good alternative to lowering the settings, but when we’ve polled our audience in the past, there are a lot of people who prefer to lower the settings than to use an upscaling approach like FSR or DLSS. It’s really going to be something that users should toggle and decide if they like. 

While that might sound like a wishy-washy answer, that is the answer. It is very situational. The good news is that it’s easy to turn these technologies on and off to see if you like it. 

Some people will prefer the higher graphics quality with the potential image quality losses that upscaling technologies provide and other people won’t. 

Going back to whether it’s worth turning on, the best answer we can provide is that we think it’s a good alternative, just like DLSS, to lowering settings in some games compared to running games at lower performance. So if you’re trying to fix performance issues and you want the higher settings, generally speaking, both the newest versions of DLSS and FSR work well for that. 

In the games we tested using FSR 4.0 as an override was an improvement over past versions, which is good. 

While AMD has made improvements to these technologies, the question you should ask is whether you want to use them at all. With frame generation, you still get some really nasty images that aren’t representative of how the game was intended to look. 

You get ghosting, doubling up of images, and distortion of text. A lot of people will see that and want to turn it off. This will also depend on the game. 

It’s a little less clear what the decision should be on the upscaling front because it will depend on how much you’re struggling to run a particular game. It becomes a choice of sacrifice between graphics settings or image quality crispness. For some people, it may not be a choice if the game is running too poorly. Overall, FSR 4, just like DLSS 4, is an improvement over its preceding version. This is what we want to see. But like DLSS, the degree of improvement on the scale of a little bit better vs totally transformative, it’s going to be closer to the a-little-better side. It won’t change how you fundamentally see games. 

Next, we may look at FSR 4 vs DLSS vs XeSS in a future content piece. That will let us look at the direct comparison between competitors.

We benchmark the RTX 5060 Ti against the RTX 4060 Ti 8GB, the RTX 3070 Ti, 3080, RX 9070, and dozens of other GPUs

The Highlights

• NVIDIA’s RTX 5060 Ti ships in either 8GB or 16GB variants
• Counter to NVIDIA’s claims, the 5060 Ti does not offer a “50x” performance increase over a GTX 1060
• The 5060 Ti is about 13%-27% better than the 4060 Ti at 1440p
• Original MSRP: $430
• Release Date: April 16, 2025

Table of Contents

• AutoTOC

Intro

The shortest possible conclusion upfront is that the 5060 Ti is about 13%-27% better than the 4060 Ti at 1440p, typically in the range of 20-25%. At 1080p, the new card is 11-24% better, typically about 18-20%. 
Against the 3060 Ti from 5 years ago, the 5060 Ti at 1440p is 16-39% improved, depending on the game. 1080p posted 21-40% gains, with a huge exception in Black Myth with ray tracing enabled, where there was a 56% uplift. The 3060 Ti was also the card that the 4060 Ti sometimes lost against.

For older devices or possible used candidates, the closest alternatives (by performance) to pay attention to in our charts will be the 3080 (watch our review) and 3070 Ti (watch our review), which often flank the 5060 Ti, and the 7700 XT (read our review) or 7800 XT (read our review) on AMD's side.

Editor's note: This was originally published on April 16, 2025 as a video. This content has been adapted to written format for this article and is unchanged from the original publication.

Credits

Test Lead, Host, Writing

Steve Burke

Testing, Editing

Mike Gaglione

Writing

Jeremy Clayton

Camera, Video Editing

Vitalii Makhnovets

Camera

Andrew Coleman

Writing, Web Editing

Jimmy Thang

But pricing is the big challenge today. NVIDIA says that this card has an MSRP of $430, with the 8GB variant at $380 and RTX 5060 non-Ti at $300, launching in May. The 5060 Ti cards launch today with the reviews. We only have the 16GB model right now. We might look at the 8GB version later.

Full transparency up-front. We’re keeping this review as simple and focused as possible, mostly because we’re currently traveling with a big story we’re working on. We still have dozens of gaming charts, but we wanted to be clear on that. 

We’re also going to keep our concluding thoughts simple because we need to see how the actual pricing shakes-out before making firmer judgments, which means there’ll be more discussion in the coming weeks -- likely in HW News or potentially another dedicated story.

With that out of the way, here’s a quick version of the specs:

The short version is that NVIDIA’s RTX 5060 Ti ships in either 8GB or 16GB variants. The review samples we’re aware of are the 16GB model. There shouldn’t be any difference between these beyond the memory, from what we’re told. These cards have 4608 CUDA cores, 144 TMUs, and a gacha box of ROPs. Memory bandwidth is rated at 448GB/s with a memory bus of 128-bit, which is why we have the multiples of 8GB for memory.

The RTX 5070 technically has a lower memory capacity at 12GB. Theoretically, they could do a 24GB model, but these options stem from the bus width and controller choices.

The RTX 5060 non-Ti will ship in May and have 3840 CUDA cores with an 8GB framebuffer, also on a 128-bit bus.

AMD’s competition will include the, in theory, RX 9060 series, for which we don’t have full details yet. We’ll hear about that more likely next month.

We’re keeping it simple today, so let’s just get into the benchmarks.

RTX 5060 Ti Gaming Benchmarks

FFXIV 4K

Final Fantasy 14 at 4K is up first.

This one was bad for the RTX 4060 Ti, with the card landing at an abysmal 41 FPS AVG as compared to the 3060 Ti FTW3’s 48 FPS AVG. RIP EVGA. We explained this regression in our 4060 Ti review previously, which we titled “Do Not Buy.” Spoiler alert: The conclusion was to not buy it.

The 5060 Ti isn’t really competing with the 4060 Ti here: It’s competing with the 3060 Ti, and against that, we see an uplift of 12% to 54 FPS AVG. The improvement over the 4060 Ti looks more impressive, but that’s because the 4060 Ti sucks. The uplift over its 41 FPS AVG was 31%. The RTX 5070’s 78 FPS AVG has it about 43% ahead of the 5060 Ti.

Used RTX 3070 (watch our review) and 3070 Ti cards might be worth exploring: We saw completed and sold listings on eBay ranging from $270 to $340, which would put them below the 5060 Ti if it even hits its marketed MSRP, which it probably won’t. The 3070 is about equal to the 5060 Ti, with the 3070 Ti slightly ahead.

AMD’s RX 7800 XT is its closest performer to the 5060 Ti, landing at 58 FPS AVG and leading the 5060 Ti by almost 7%. 

AMD’s RX 7600 (watch our review) falls way down the chart and runs at 32 FPS AVG. But then again, AMD does overall poorly in this particular game, with its 9070 XT (read our review) down below the RTX 5070. We talked about that in our 9070 series reviews.

Finally, NVIDIA’s claimed 50x performance increase over the GTX 1060 (watch our revisit) doesn’t come to fruition when not arbitrarily enabling and disabling favorable settings. The 1060 ran at 16 FPS AVG. 16 x 50 is 800 FPS, which would be 4x the performance of an RTX 5090 (read our review). We’ve gone beyond an RTX 5070 = 4090 and to a RTX 5060 Ti = RTX 9090. In reality, the 5060 Ti is 236% ahead. That’s still a big jump, but no need to stretch the truth about it.

FFXIV 1440p

At 1440p, the RTX 5060 Ti landed at 104 FPS AVG, which has it functionally tied with the 3070 Ti’s 108 FPS AVG and only slightly ahead of the 7700 XT’s 98 FPS AVG or 3070’s 97 FPS AVG. The lead over the 3060 Ti is now 16%, or about 25% ahead of the 4060 Ti. These gains are down from 4K. Lows are where you’d expect them for each card, with no meaningful differences.

The 5070 is about 47% ahead of the 5060 Ti, slightly up from the lead at 4K.

AMD’s 9070 ran at 126 FPS AVG here, producing a 22% advantage. The MSRP is higher, but then so is everything, including the companies that set these prices.

The GTX 1060 ran at 33 FPS AVG here, the 1650 (watch our review) at 23 FPS, 6500 XT (watch our review) at 31 FPS, and 3050 (watch our review) at 45 FPS AVG. Predictably, the 5060 Ti is a big improvement over all of these, but again, not 50x.

FFXIV 1080p

At 1080p, the 5060 Ti ran at 160 FPS AVG. That’s about the same as the 7700 XT and slightly ahead of the 3070 Ti. It’s finally moving up the relative ranking compared to Ampere. We think you’d still be better off with a used card right now. 

The 4060 Ti ran at 133 FPS AVG here, so the 5060 Ti improves by 20%. Against the 3060 Ti, the 5060 Ti is about 25% better. The 4060 Ti is finally better than the 3060 Ti when at 1080p, so those two have flipped as well.

The lower-end round-up includes the GTX 1060 at 51 FPS AVG, 3050 at 66, 1070 (watch our review) at 70, and 2060 (watch our review) at 83. The 6600 (watch our review) ran at 86 FPS AVG and AMD’s 7600 ran at 107 FPS AVG.

Black Myth: Wukong - 4K

Black Myth: Wukong is up now, first at 4K. The 5060 Ti ran at 31 FPS AVG. That’s obviously unplayable and is because of the resolution and settings, but it’s still useful for relative scaling.

The result has it about equal to a 7800 XT. The 3080 leads by 18%, with the 5070 leading by 29%. The 3070 Ti ran a lower framerate than the 5060 Ti in this one.

Let’s move to something more playable.

Black Myth: Wukong - 1440p

At 1440p, the 5060 Ti ran at 57 FPS AVG and landed right between the 7800 XT and RTX 4070 (watch our review). These are all effectively tied. The 5070 held a 72 FPS AVG, or 27% ahead. That’s slightly down from 4K. The 5060 Ti is ahead of the 3070 Ti by 13%, the 4060 Ti by 27%, and the 3060 Ti by 34%. This is one of the games where the 4060 Ti and 3060 Ti are right next to each other. 

AMD’s 7600 ran at 31.6 FPS AVG and isn’t really in the same class of card as what we’re reviewing today. Its 7800 XT and 7900 GRE (read our review) are the most comparable, but in theory, the inbound 9060 XT should be fighting in this territory.

Black Myth: Wukong - 1080p

At 1080p, the 5060 Ti pushed 81 FPS average with lows where you’d expect given the average. There is no particularly special frametime consistency benefit.

The 5060 Ti ends up at about the same level as the RTX 3080 and RTX 4070 (watch our review). The 5070, which still doesn’t equal a 4090 (no matter what NVIDIA says), and its 98 FPS AVG puts it 21% ahead of the 5060 Ti.

As for the last generations, NVIDIA’s new 5060 Ti leads the 4060 Ti by 24%, or the 3060 Ti by 40%. 

AMD’s 7900 GRE is its closest card here, slightly ahead of the 5060 Ti, with the 7800 XT just behind. 

Intel’s B580 cards are at around 46 FPS AVG, which has them similar to the RX 7600.

Starfield - 4K

In Starfield at 4K, the 5060 Ti ran at 41 FPS AVG with lows at 34 and 29, proportional to the cards around it. This puts the 3080 ahead of the 5060 Ti and the 5060 Ti ahead of the 3070 Ti. 

AMD’s 7800 XT outdoes the 5060 Ti by 18%, with the more expensive 9070 non-XT up at 63 FPS AVG.

The RTX 5070 ran at 54 FPS AVG here, 32% ahead of the 5060 Ti. We’ll move to 1440p for prior generations.

Starfield - 1440p

At 1440p, the 5060 Ti landed at 65 FPS AVG. Unfortunately for NVIDIA, this is a terrible result: The 4060 Ti was at 58 FPS AVG, narrowing the uplift to only 13.4%. That isn’t a big improvement. The lead over the 3060 Ti’s 50 FPS AVG is 30%, also not that impressive for two generations.

AMD’s 7800 XT leads the 5060 Ti by 15.5%, with the 9070 (read our review) obviously way ahead given its higher theoretical price and positioning.

For those considering used options, the 5060 Ti only outdoes the 3070 Ti by about 8%, making it a reasonable alternative that might save some money.

Starfield - 1080p

At 1080p, the 5060 Ti held an 82 FPS AVG with lows positioned about the same as everything around it. The 3070 Ti’s 76 FPS AVG encroaches on the 5060 Ti’s result and, from an actual human perspective, would look about the same. The 4070 outperforms the 5060 Ti by 16%, with the 5070 ahead by 26%. 

The 4060 Ti’s 74 FPS AVG means the 5060 Ti is about 11% better, overall a boring generational jaunt. The uplift over the 3060 Ti is 28%.

Dragon’s Dogma 2 - 4K

Dragon’s Dogma 2 at 4K is up next.

The 5060 Ti ran at 40 FPS AVG, so it’s nearly exactly tied with the 3070 Ti. The 0.2 FPS AVG advantage is well within run-to-run variance. Lows are also tied. AMD’s 7800 XT is 17% ahead of the 5060 Ti’s average FPS, with the 5070 ahead by 41%.

We’ll move to lower resolutions to look at the prior generation 60 and 60 Ti-class cards.

Dragon’s Dogma 2 - 1440p

At 1440p, the 5060 Ti ran at 70 FPS AVG, planting it right in the middle of the 3080 and 7700 XT. The 3070 Ti is right behind with a 65 FPS AVG, with the 4060 Ti at about the level of the 3070. The new 5060 Ti leads the 4060 Ti by 22% and the 3060 Ti with its 53 FPS AVG by 32%.

AMD’s 7600 is far down this chart, so it’ll need something newer in the 9060 class to compete here. Intel’s B580 is also down near the RTX 4060 and RX 6600 XT (watch our review).

As for what’s better than the 5060 Ti: Other than the 3080, the 4070 is about 12% better and 5070 is about 37% higher average FPS.

Dragon’s Dogma 2 - 1080p

At 1080p, the 5060 Ti ran at 93 FPS AVG, landing between the 3080 and 7700 XT again. The improvement over the 4060 Ti is just 19%, followed by the 3060 Ti’s 68 FPS AVG for an uplift of 36%. The 3070 Ti gives the 5060 Ti just a 10% lead, doing better than the 4060 Ti.

As for the GTX 1060, considering 50x its performance would put it over 1,100 FPS, we’d say NVIDIA missed the mark on this by orders of magnitude.

Cyberpunk 2077: Phantom Liberty - 4K

Cyberpunk 2077: Phantom Liberty is next. This is newer data, so we haven’t re-run the 4060 Ti, 4060, and 3060 series cards through here yet. We’ll show it anyway for the other comparisons.

At 4K/Ultra without RT first, the RTX 5060 Ti ran at 31 FPS AVG. This has it meaningfully ahead of the 3070 Ti by percentage, improved by 19%. The 7800 XT leads the 5060 Ti by about 10% here, with the 5070 about 30% ahead of the 5060 Ti.

Cyberpunk 2077: Phantom Liberty - 1440p

At 1440p, the 5060 Ti ran at 68 FPS AVG, putting the 5060 Ti between the 3080 and 3070 Ti. The 5070 ends up about 30% ahead with its 88 FPS AVG, meaning that, if we just pretend that the MSRP numbers stick, you’re paying about 1% more money for every 1% more performance between the 16GB 5060 Ti and 5070. 

Cyberpunk 2077: Phantom Liberty - 1080p

At 1080p, the 5070 is down to a 27% lead over the 5060 Ti. The 5060 Ti now leads the 3070 Ti by 20% and the 3070 non-Ti by 27%. The 7800 XT is about 8% ahead of the 5060 Ti here.

Let’s move to something where we have last-gen numbers.

Dying Light 2 - 1440p

Dying Light 2 at 1440p is one of the situations that was bad for the 4060 Ti versus the 3060 Ti: The two cards are indistinguishable, with performance identical between them. The RTX 5060 Ti ran at 74 FPS AVG, so it outperforms the 4060 Ti (and therefore 3060 Ti) by about 23%. It took them two generations, but they’ve finally beaten the 3060 Ti in this game. The 4070 is about 6% better than the 5060 Ti here, with the 7800 XT about 14% ahead. The 5070 leads the 5060 Ti by 44% here, so if anything, the 5060 Ti stands to make the 5070 look better. Against the Intel B580’s 63 FPS AVG, NVIDIA’s 5060 Ti is about 17% better in one of the better B580 showings. 

Dying Light 2 - 1080p

1080p has the 3060 Ti and 4060 Ti again roughly adjacent to one another, with the 5060 Ti leading the 4060 Ti by 19%. The 5060 Ti’s lead has diminished from the 1440p result. The 5060 Ti is similar to the 7700 XT’s performance here, including in lows, with the 4070 leading the 5060 Ti by almost 9%.

Against older cards, the 5060 Ti improves on the GTX 1060 by not 50x, to nobody’s surprise, and instead by about 3.8x. Even with MFG, you would not get 50x. Maybe with DLSS at Ultra Sh*t quality and upscaling from 144p, we’re not certain you could squeeze 50x out of this lemon, though. You’d have to go out of your way to hurt the 1060.

For those still on an RTX 2060, you can expect about a doubling of performance to the 5060 Ti in a scenario like this. 

Resident Evil 4 - 4K

Resident Evil 4 is up now, first at 4K and without ray tracing. The RTX 5060 Ti ends up performing about the same as the 7700 XT. The RTX 4070 leads the 5060 Ti by about 10-11% here, at 63 FPS AVG to 57, with the 5070 leading by 38%. That’s similar to what we’ve seen elsewhere so far. The 3060 Ti also launched for $400. Adjusted for inflation, that amounts to $491. The new card is $430. So things haven’t changed that much. The price is similar/slightly lower and performance has hardly improved. The improvement over the 2060 is about 135%. 

Resident Evil 4 - 1440p

At 1440p, the 5060 Ti leads the 3070 Ti by 13% and the 4060 Ti by 23%, followed by the 3060 Ti at 36.5%. The reduced resolution has benefitted the 4060 Ti marginally, allowing it to distance itself from the 3060 Ti. The B580 (read our review) is actually around the 3060 Ti’s performance, excepting 1% lows.

As for the RTX 5070, which remains not a 4090 (watch our review), NVIDIA’s biggest lie leads the 5060 Ti by 36%. 

AMD still doesn’t have a new and direct competitor here, but probably will in May in the 9060 series. For now, the 7700 XT is the closest and outperforms the 5060 Ti slightly.

RTX 5060 Ti Ray Tracing Benchmarks

Ray tracing is up next. These games are generally a heavier load, so we have a mix of upscaled benchmarks and of native resolution benchmarks, but all of them are with RT on now.

Ray Tracing - Black Myth: Wukong 4K

In Black Myth: Wukong at 4K upscaled, the RTX 5060 Ti landed at 29 FPS AVG. That has it ahead of the RX 9070, which is more of a problem for AMD than it is a positive for NVIDIA. We already knew this about this game and AMD, though. The 9070 XT ends up about tied with the 5060 Ti here, and actually, the 3090 non-Ti (watch our review) at 29.8 FPS AVG is only about 3-4% ahead of the 5060 Ti. This game remains heavily favored for NVIDIA, especially with ray tracing. 

The RTX 5070 has a large lead at 40 FPS AVG, although its memory capacity can prove problematic in some of these heavier scenarios. The 5070 leads by 39% here. The 5060 Ti also shows more meaningful gains over the 3070 Ti in this test than in some of the raster tests, at 31% improved.

Ray Tracing - Black Myth: Wukong 1080p

At 1080p upscaled, the 5060 Ti's 74 FPS AVG has it ahead of the 9070 XT, 3090, and 3080. It also leads the 4060 Ti by 19% and the 3060 Ti's 48 FPS AVG by 57%.

The generational RT uplift is helping the 5060 Ti distinguish itself more here than it did when rasterized.

Against the first-gen ray tracing 60-class card, the RTX 2060, we're seeing a 175% improvement.

Ray Tracing - Dragon’s Dogma 2 4K

In Dragon's Dogma 2 at 4K and with ray tracing, the RTX 5060 Ti ran at 35 FPS AVG and roughly tied (but technically led) the 3070 Ti. That has it better than the 2080 and 2080 Ti of years past, although the 3080 still manages to best the 5060 Ti. AMD's 7900 GRE outperforms the 5060 Ti slightly, with its newer 9070 cards performing up around 3090 Ti levels -- but at a higher theoretical base MSRP than the 5060 Ti.

Ray Tracing - Dragon’s Dogma 2 1440p

At 1440p with RT, the 5060 Ti ran at 62 FPS AVG and kept the lows consistent with the average. There's nothing particularly impressive or bad for the frametime consistency and lows. It’s just kind of where we’d expect it. 

The 4070 and 7800 XT are about 9% ahead of the 5060 Ti. And for that matter, the 3080 is around that same area. The 5070's 82.8 FPS AVG is around 34% ahead of the 62 FPS for the 5060 Ti. We've seen higher in other games.

As for the lower-rank cards, the 4060 Ti ran at 49 FPS AVG and yields a 26% uplift to the 5060 Ti. The 3060 Ti isn't far behind the 4060 Ti in this one, but at least they're in the order you'd expect them to be. 

Ray Tracing - Dragon’s Dogma 2 1080p

At 1080p, the 5060 Ti's 82 FPS AVG landed it between the 3080 and 7700 XT again. This has been consistent. This result gives it a 20% improvement over the 4060 Ti and 37% improvement over the 3060 Ti.

Ray Tracing - Dying Light 2 4K

Dying Light 2 is up next. Dying Light 2 at 4K upscaled with ray tracing has the 5060 Ti at 31.8 FPS AVG and exactly tied with the 7900 GRE for average and 1% lows. The 5070 improves to 43.7 FPS AVG, or 37% once again. This seems to be a fairly consistent percentage improvement to the 5070.

Ray Tracing - Dying Light 2 1440p

At 1440p upscaled with RT, the 5060 Ti ran at 60 FPS AVG. We haven't yet re-run the 4060 Ti or 3060 Ti in this one, leaving us to compare instead with the 7900 GRE -- where they're about equal -- and the 3080, which remains a bit ahead of the 5060 Ti. The 5070 leads at 81 FPS AVG, or 34%.

Ray Tracing - Dying Light 2 1080p

At 1080p, we re-introduce the 4060 Ti and 3060 Ti. The 5060 Ti's 87 FPS AVG has it about 20% ahead of the 4060 Ti, which itself was only 11% ahead of the 3060 Ti. The B580 is actually somewhat close here, roughly tying the 3060 Ti. As for AMD, the 7900 GRE remains the next closest to the 5060 Ti.

Ray Tracing - Resident Evil 4 4K

Resident Evil 4 is up with ray tracing now, first with 4K and upscaled. The 5060 Ti's 67 FPS AVG has it tied with the 7700 XT, including in the 1% and 0.1% lows. The 4070 and 3080 lead these results, as they have for the past games. The 5070 leads the 5060 Ti by 36%, close to prior results. As for the 4060 Ti, its 52 FPS AVG gives the 5060 Ti a lead of 30% for one of the larger gaps.

Ray Tracing - Resident Evil 4 1440p

At 1440p, the 5060 Ti again falls between the 7700 XT and RTX 3080. The lead over the 4060 Ti is narrowed to 24% now, with the uplift over the 3060 Ti at 39%.

Ray Tracing - Cyberpunk 1080p RT Medium

In Cyberpunk 2077: Phantom Liberty at 1080p RT Medium, the 5060 Ti ran at 63 FPS AVG and sat between the 7900 GRE and 7900 XT. The lead over the 3070 Ti is about 10 FPS AVG here, or 17%. The 5070's 82 FPS AVG has it 31% ahead of the 5060 Ti, down in relative improvement from other benchmarks.

Ray Tracing - Cyberpunk 1080p RT Ultra

At 1080p and with RT Ultra, the 5060 Ti ran at 50 FPS AVG (which is really not bad when considering how heavy this workload is), or just ahead of the 7900 XT Hellhound. The 3070 Ti hit 40 FPS AVG here, with lows suffering for the 0.1% value. The 5060 Ti's lows are OK in this one, supported by the 16GB capacity.

Conclusion

We’ve provided the benchmark numbers above. At the very least, you have the data you need to figure out if an upgrade makes sense for you. This is going to be one where we withhold full value judgment until it properly launches because we do not trust the MSRP to persist for the majority of purchasers.

It certainly isn’t going to 50x a GTX 1060, though.

GPU Price Comparison | GamersNexusApril 2025

The 16GB RTX 5060 Ti’s MSRP is set at $430, whether or not we see it at that. That’s better than the previous 16GB 4060 Ti’s launch MSRP of $500, which was atrocious. The 4060 Ti 8GB model was $400, so 8GB more memory used to be a $100 upsell, but no one bought that, so now it’s a $50 upcharge. 

Accounting for inflation favors the 5060 Ti over the 4060 Ti for the 16GB models with MSRPs. 

Nearest performance neighbors to the 16GB RTX 5060 Ti are typically the RTX 3080 and RX 7800 XT above, and RX 7700 XT and RTX 3070 Ti commonly below.

We’d love to dig into value comparisons between the 5060 Ti and its current competitors, but the availability of GPUs at retail in this price bracket ($380-$480) is terrible. 

Using Newegg as a representation shows only 9 SKUs of any video card sold by Newegg in stock in that price range, and only one of them is actually brand new. It’s an EVGA RTX 3060 XC (not a Ti) for $440, which is an awful price.

So without any new cards to buy in this price bracket, we can look toward used options -- and this is where we think people should be seriously looking.

The RX 7900 GRE was around $627 average for sold listings, followed by the RX 6950 XT and 7800 XT at $556-$575 on average and the RX 7700 XT at $472 average.

There’s a potential edge case where a good deal on an RX 6950 XT could be an interesting higher-performing wild card – but that’s only if you’re entirely focused on raster performance, and you’re willing to hunt for a deal on one around $500. Higher power consumption is also something to consider. NVIDIA’s 5060 Ti is more efficient.

A used RTX 3070 Ti at $358 average would be a good lower-performing budget option below the MSRP of a new RTX 5060 Ti of any capacity. We found some that were in the upper $280-$290 to lower $310 range, which would be worth seriously considering. VRAM may be limiting in some situations. The RTX 3080 is going for around $449 and typically beats or is close to the 5060 Ti. The 3080 isn’t cheap enough on the second-hand market yet to get our strong recommendation in this specific scenario. That’s doubly the case for used RTX 4070s at the time of writing, which are newer and have been selling for $621 on average.

And that brings us to what we’ve said a lot in the past: if your computer is doing what you need it to do and if you don’t feel a need to upgrade, then we’d say hold off. But some people do either “need” to buy new devices to replace aging hardware or just really want the escape of building a PC, which we also appreciate and relate to. It’s just going to come down to your price tolerance.

Right now, we really don’t have the answers. 

We’re trying to figure them out, which is why we’re currently flying all around the US at our own expense to talk to companies about the pricing situation. We hate to not be able to give a value judgment at the end of a review, but until this card is actually available -- which will coincide with the launch of this review -- we just can’t know what it’s going to cost.

Once it is, we’re going to do a recap either in the news or standalone.

Just for context: We just met with a distributor that buys hundreds of thousands of GPUs per year. Last week, we saw their cost to buy RTX 5090s. The cheapest was around $2,400 up to $3,000, and that’s their cost. That means $2,000 is impossible. We’re not sure to what extent that’ll affect the 5060 Ti cards, especially after the launch period where pressure to maintain the price is off.

Now maybe that’ll change with the partial tariff exemptions, but those are up in the air.

We investigate whether NVIDIA’s new drivers are behind the recent GPU instability issues

The Highlights

• We read through numerous user reports and set forth to replicate NVIDIA’s stability problems
• NVIDIA has major driver instability problems right now, and that impacts newer RTX 50 series cards as well as RTX 40 and 30 series GPUs
• We found issues with setups using DLSS4, frame generation, G-Sync, and even just the monitor ports in a multi-monitor setup

Table of Contents

• AutoTOC

Intro

How the tables have turned. NVIDIA, is now playing the role of AMD circa Vega, which is not good. NVIDIA’s drivers have had stability problems since the launch of its RTX 50 series GPUs and we’ve been able to replicate them.
In one game, we had system crashes after enabling NVIDIA’s frame generation, which is unfortunate since it marketed the 50-series on the back of MFG and fake frames. We had reboots loading into Cyberpunk, crashes and driver errors in Tomb Raider, and issues with screen distortion and artifacting.

Editor's note: This was originally published on April 6, 2025 as a video. This content has been adapted to written format for this article and is unchanged from the original publication.

Credits

Test Lead, Host, Writing

Steve Burke

Camera, Video Editing

Vitalii Makhnovets

Testing, Writing, Research

Tannen Williams

Writing, Web Editing

Jimmy Thang

But these issues didn’t show up in our review of the 50-series, which was our first hint that it’s related to a driver version issue. 

Game developers have recently begun recommending rolling back to older NVIDIA drivers: 

inZOI, the studio which NVIDIA has heavily promoted for partnership with its AI features, threw NVIDIA under the bus, stating, “Using driver versions 572.xx or later may result in occasional frame drops or stuttering [...] If issues persist, we recommend installing version 566.36.” This driver is from December. 

Developer Neople stated, “Using driver versions 572.xx or later may result in occasional frame drops or stuttering.”

This aligns with our findings from the past week, where we’ve been trying to replicate these issues that have been widely reported online. 

Interestingly, we found that the order of the monitors in Windows, as in the display out order of monitor 1 and monitor 2 in the OS via the actual hardware connection to the video card, seemed to affect stability. 

User reports commonly note various display problems like black screens, flickering, and sometimes even unprompted system reboots and crashes to desktop. This includes other types of crashes too, like crashing to desktop. Upon a closer inspection, it wasn’t only 50-series users dealing with the issues. Many users of NVIDIA’s previous generations of GPUs, particularly the 40 and 30 series, are also affected by the recent drivers. 

Several users speculated on the impact of multiple monitors, the impact of mixed or higher refresh rates, and even how NVIDIA’s own G-Sync might be affecting the driver stability. 

There’s been at least one megathread on Reddit with hundreds of reports and these issues have been in the news since the launch of the 50-series. 

We’ve been able to replicate these issues and have isolated some of the causes. We also have some stopgap solutions. They’re not great, but may help for the time being. This also applies to multiple card generations. As far as we know, it impacts the 30 to 50 series GPUs.  

User reports seem to point to driver 566.36 as the most stable recent driver. 

This version was released on December 5th, 2024, just prior to the 50 series launch in January. In the four months since that driver’s release, NVIDIA has posted six GeForce Game Ready Drivers and at least four GeForce “Hotfix” Drivers specifically aimed at fixing bugs and system crashes.

Testing Configuration

Time to get into our testing configuration.

In order to better understand the driver problems that previous generation NVIDIA users are currently experiencing, we attempted to recreate some of the issues ourselves. 

Before we began, we went through an assortment of user reports to find games, system configurations, and specific settings that seemed especially problematic to get a starting point for our testing.

For our testbench setup, we used a modified version of our standard test bench. We also bought a couple games that people had particular issues with. Our standard GPU review bench did not exhibit any of the issues with drivers initially, so we modified it.

Components included these:

• CPU: 9800X3D
• GPU: 4070 Super
• Monitors: Dual-monitor setup using an Acer 4K 60Hz monitor (not G-Sync compatible) and a Dell Alienware 1080p 360Hz monitor that is G-Sync compatible, which we ran at 240Hz. Both were connected via DisplayPort. And this part is important: In Windows, we manually set the monitor order:

• Monitor 1: Acer 4K display
• Monitor 2: Alienware monitor, which was checked as the main display in the Windows settings). 

• Motherboard: ASUS ROG X670E Hero
• RAM: 2x16GB (32) of DDR5 6,000Mhz Trident Z5
• PSU: EVGA 1600 T2
• Driver version: We used 572.83 initially as it was the most recent when we began our testing, which was released on March 18th, 2025

So we have a very intentional mix of different resolution displays, different G-Sync compatibility, different refresh rates, and, of course, we’re going to change the drivers as we go through our testing.

Our first goal was to simply identify a game crashing. From there, we’d recollect what occurred leading up to the failure and then document a process that we could follow to ensure the crash happened every time. Once we were able to consistently reproduce the complication, we’d change one variable at a time while keeping all others the same in order to isolate the factors involved and determine possible causes. For each instance, we separately disabled G-Sync, frame generation, DLSS, and sometimes other settings (depending on the game). We also moved to a single monitor and then swapped output ports on the GPU between the two monitors for each game issue we experienced. Suffice it to say, this all took a while.

Broad Findings

Here’s what we found:

In our testing on NVIDIA’s most recent driver, 572.83, we found four different instances of game crashes that could be consistently reproduced. 

In Star Wars Outlaws, we experienced failures after selecting “resume” in the startup menu. In Marvel Rivals, we’d observe system reboots either directly after applying frame generation in the pre-menu settings, or after applying the settings and then exiting to desktop from a match. 

In Cyberpunk 2077, we encountered reboots when loading into the game. 

In Shadow of the Tomb Raider, when running the in-game benchmark we faced a game crash along with a driver error. After rolling back to a previous driver, we could no longer replicate any of these same failures on any of these games. And that’s the reason this took a week or 2 to put together because we tried to isolate to see if it was a game problem, something relating to NVIDIA, or NVIDIA settings in that game.

Based on our initial findings, we believed the driver issues may be affecting some users that are operating on a specific combination of hardware and displays/graphics settings. Specifically, those using two monitors with G-Sync and frame gen enabled. We also found something really interesting regarding the physical output ports on the GPU itself and how the monitors are identified in Windows display settings. We don’t fully understand it yet, but we’ll go over that observation shortly.

We have more findings to go over first.

Less Conclusive Findings

DLSS and Reflex may also factor into the equation, but unfortunately we weren’t able to make a clear determination. In some games, DLSS or Reflex is automatically enabled when using frame gen. Meaning, we weren’t always able to fully isolate each of these variables because one will force the other. Additionally, sometimes lowering the refresh rate to 120Hz on our second monitor would resolve issues, while other times it didn’t. 

Something that especially stood out to us was that we could only replicate 3 of the 4 failures when both monitors were connected via DisplayPort and when monitor 2 was checked as the main display in Windows settings. 

As seen in the image above, when our Acer monitor was plugged into what we’re calling slot 3 and our Alienware monitor was plugged into what we’re calling slot 2, we didn’t have any issues. But when we swapped the slots the monitors plugged into, effectively converting “monitor 1” to “monitor 2” within Windows and vice versa, and then made monitor 2 the main display, we saw the crashes again consistently. 

This is less to do with the physical ports being used and more to do with how Windows and the drivers are treating the monitors in their “1” and “2” slots. We couldn’t tell you why this behavior causes issues, but if you’re having trouble, swapping monitors 1 and 2 could potentially help.

We actually found this by accident. The day after we recorded our first game crash, we reconfigured the bench and began trying to trigger it again, but couldn’t. It wasn’t until we realized that the only difference was that the DP output slots on the GPU had been changed. After switching them back to the original position, we were suddenly able to observe the failure again. To our surprise, this same circumstance appeared in multiple games.

Game Crashes

As for similarities between the issues: In Star Wars Outlaws, Marvel Rivals, and Cyberpunk 2077 (all games that natively support DLSS 4), we were unable to produce a failure when either G-Sync or frame gen were disabled. We also separately couldn’t trigger a crash after swapping output slots or on a single monitor setup.

DLSS 4 and frame gen are also problems here. It’s not just the monitor order or the monitor ports.

The error in Shadow of the Tomb Raider was different from the games previously discussed. This game was unaffected by G-Sync. It also doesn’t feature a frame gen setting (at least the version we tested doesn’t). Additionally, it experienced the same failures on a single monitor and after swapping output slots. So none of the other issues we’ve talked about thus far were the cause for Shadow of the Tomb Raider.

Instead, we were able to boil this problem down to having “RTX Shadow Quality” set to Ultra in the in-game settings. And we know for a fact that this didn’t always cause a crash because we have run tests on it with RTX Shadow Quality set to Ultra in prior versions of our benchmark suite with different drivers. 

In regards to our analysis, the main commonality between all four of these games is that they crashed on NVIDIA’s most recent drivers at the time of testing, but after rolling back to 566.36, we were no longer able to trigger that same error. 

Rolling back to driver version 566.36 doesn’t guarantee that crashes will stop. From what we’ve read online, there are still issues with that driver. In our case, we are unable to prompt the same failures we had previously been able to. This, along with various user reports, seems to suggest that driver 566.36 is currently the most stable driver that NVIDIA has released recently. 566.36 is from December and predates the launch of the RTX 50 series, so would not be compatible with 50-series cards.

There are a lot of other issues reported online, but that’s what we’ve replicated so far.

Many users saw problems in games that didn’t occur until after multiple hours of gameplay. As much as we’d like to, we can’t spend several hours playing a random game to see if it will possibly crash. It’s just too unpredictable. On the same note, just because we weren’t able to repeat our failure when we turned G-Sync or frame gen off, doesn’t mean that turning these options off will solve all problems, but hopefully it’ll help some of you.

Other Common Issues

There were some other issues we commonly saw in our research:

We’ve seen multiple reports that there’s an issue with waking from sleep or resuming from idle states and we were able to create at least one of these by accident. We noticed on one of our personal PCs that there were issues waking the system from sleep when the PC had been idle. And on wake, the monitors appear to turn on but don’t receive any signal and the screens remain black. Fixing this required a PC reboot. Windows has its own decades of issues with sleep and S3 states, so it’s hard to specifically pinpoint who’s at fault for this one between NVIDIA and Microsoft.

Another issue we’ve seen going around has to do with the alt+tab shortcut acting buggy and freezing games when switching windows while playing a game. 

We did occasionally encounter this bug, but it wasn’t something we saw every time and seemed to be resolved pretty easily by pressing the Windows key or alt+tabbing back into the game. Again, this one is difficult to pinpoint between NVIDIA and Microsoft. 

NVIDIA Driver Notes

NVIDIA has had some driver notes addressing parts of this, despite the driver not fixing everything.

In NVIDIA’s 572.83 driver release notes, one open issue states, “[Cyberpunk 2077/Half-Life 2 RTX] PC may bugcheck with error 0xd1 when playing game while using DLSS Frame Gen + G-SYNC [5144337].” 

We’re glad that issues relating to frame gen and G-Sync are being acknowledged, but we’d like to see more from NVIDIA here. This is a relatively vague description and only mentions two games that are affected by this much larger issue. 

NVIDIA has also had multiple hot fixes. We’ve reported on some of them where it’s saying it’s trying to fix issues with the original 50 series and then just basically everything. 

Suggestions for Those Affected

We’ve known that NVIDIA has shifted its focus towards larger business clients and away from its gaming market. That’s fine -- but it shouldn’t be screwing up areas where it has long had such refinement. NVIDIA is known for its driver stability, and the company, in its series of screwups for 2025, is starting to chip away at that in an era where AMD’s drivers have become relatively stable. These driver issues are another mark upon NVIDIA’s tarnished gaming reputation with the 50-series launch as the company shifts focus, but losing prior-gained ground shouldn’t be in the cards, and NVIDIA hopefully will recognize that these issues are damaging to its long-term brand credibility.

We’ll end off with some suggestions for users experiencing issues:

For non-50 series users who are confronted with game crashes on NVIDIA’s latest drivers, we’d first recommend swapping the order of your display outputs on the GPU itself. We don’t fully understand how effective this fix is yet, but it’s simple and worked for us. We’d also suggest swapping cables to HDMI, which is not ideal, to see if that changes anything as a stopgap.

If you don’t feel strongly about G-Sync or frame generation being enabled, we would then suggest disabling one of those settings and see if you’re able to resolve the issue that way until NVIDIA is able to get its s*** together. Considering NVIDIA’s whole marketing gimmick lately has been frame generation, this is a bad solution for NVIDIA, but hopefully it’s a stopgap.

If neither of the above work for you, our last suggestion would be to rollback to a previous driver. If you’re able to recall a specific driver that you previously used that didn’t cause any problems for you, we’d advise reinstalling that one. 

In our case, 566.36 seemed to remedy our problems and seems to be shared by users as one of the more stable ones currently available. 

As general process control, we’d recommend installing drivers directly from NVIDIA’s website and disconnecting from the internet for installs and uninstalls. We always uninstall previous drivers using Display Driver Uninstaller (DDU) in safe mode, pause windows updates, and perform a clean install before reconnecting to the internet. It’s not something that we experienced, but we have seen user reports claiming that installing drivers using the NVIDIA App led to certain issues, and reinstalling the same driver but from NVIDIA’s website instead solved those problems. 

Conclusion

Unfortunately, all of these suggestions are only temporary workarounds, and we can only wait for a new NVIDIA driver to provide a legitimate solution and get it together. This has been an absolutely abhorrent and completely embarrassing launch for the company. This is the worst NVIDIA launch we’ve covered and it may be the worst GPU series launch we’ve ever covered. That’s saying a lot considering AMD has had some impressively bad drivers for years. NVIDIA has taken that over. At least with Intel Arc and its Alchemist GPUs, our expectations were low, and Intel has taken notable steps to address its issues. NVIDIA, on the other hand, continues to fumble in impressively bad ways. This is the type of thing that helps the company lose market share. 

In the meantime, we will continue to monitor the driver situation and will provide an update when we find a more stable solution provided by NVIDIA.

We look at how NVIDIA has downsized essentially all of its gaming GPUs in terms of relative configuration compared to each generation’s flagship

The Highlights

• This article expands upon our "RTX 4080 problem" by looking at the entirety of the RTX 50 series, including how the RTX 5070 looks an awful lot like a prior 50-class or 60-class GPU.
• NVIDIA is giving you the least amount of CUDA cores for a given class of GPU than ever before.
• GPU prices have crept higher across the board, but NVIDIA's, in particular, have lost step with what we came to expect from generations of GPU launches.

Table of Contents

• AutoTOC

Intro

NVIDIA is giving you the least amount of CUDA cores for a given class of GPU than ever before. 

Today an RTX 5070 is comparable to a GTX 950 (watch our review) in some ways when you run some numbers. An RTX 5080 isn’t distant from a 2060 (read our review) in some considerations. The relationship between the number of CUDA cores the flagship has and the number of CUDA cores the lower-tier GPUs has been getting worse basically across the board. The amount of money you have to spend, even adjusted for inflation, to buy the GPUs has been staying flat or rising.
When this happens in any other product category it’s called shrinkflation.

Editor's note: This was originally published on April 3, 2025 as a video. This content has been adapted to written format for this article and is unchanged from the original publication.

Credits

Host, Writing

Steve Burke

Writing, Research

Jeremy Clayton

Video Editing

Vitalii Makhnovets

Writing, Web Editing

Jimmy Thang

Chart - Percentage of CUDA Cores Relative to Generational Flagship

We’ve already talked about the pricing issues and availability, but what we’re doing now is revisiting a topic that we ran about 2 years ago in a video called the RTX 4080 problem, in which we explored why no one was buying 4080s (watch our review) at the time. It wasn’t just the money but because the relationship of what you got for the money. We’re taking the concepts where we broke out the pricing, the components, the die area, etc. and applying it to the 50 series and, in short, it has not gotten better.  

We just have 2 main charts to go through in this article but they’re really interesting. Now that NVIDIA has shipped everything except for the 60-class card, we’ve got a good amount to look at. The real goal of this is to explore the relationship between the money and what you get for it, but we’re also going to compare some of the cards against prior generations and doing some inflation adjustments.  

We started working on this piece for the 5080 launch and then realized it’s going to get worse. So, we waited for the 5070, which is now here (unfortunately). Let’s get into the data for this.

This chart compares the percentage of the Flagship CUDA core count that each configuration is. Due to architectural changes, we’re not interested in the raw count of CUDA cores, but the percentage occupancy of the maximum config for the flagship die.

Our chart plot tracks each same-named GPU class across NVIDIA generations on a percentage scale representing how many CUDA cores each has relative to a larger configuration. The GPUs are all shown relative to the CUDA core count of that generation’s top gaming, non-Titan GPU – the 5090, 4090, 2080 Ti (watch our review), 1080 Ti (read our revisit), and so on – which we’re calling the Flagship-class. If you see 100% anywhere, that means it is equal in CUDA core count to the flagship.

We went with the 3090 (watch our review) for the 30 series rather than the late-arriving, cash-grab, full die 3090 Ti (watch our review). We had to make a judgment call.

The Flagship-class is plotted relative to the largest die’s maximum possible core count. The GTX 780 Ti (watch our revisit) is one of the few exceptions where NVIDIA made a flagship with the full non-cut-down die. 

The GTX 780 (watch our revisit) had 80% of the CUDA cores that the flagship 780 Ti did. The RTX 2080 (watch our review) brought that down to 68% of full CUDA count, then just 59% for the 4080, but it gets worse. The 5080 is a mockery of an 80-class card with only 49% of the flagship-class CUDA count configuration. We don’t care if the die is different or not for this chart, just the config.

The 3080 temporarily bucked the trend at 83%, which was great. This correlates with its incredibly good value and performance at launch with positive reviews. Back in our review of the 3080, we said, “The card performance overall is impressive. It’s a big recovery from the 20 series when we reviewed it and called 3 of the cards a waste of our time because they were 1080 Tis and then complained for 55 days about how there was no RTX and the cards were named RTX. So this was a big turnaround for NVIDIA.”

That also, however, aligns with the reviews we and others gave to the 3090 and 3090 Ti. For example, in our 3090 Ti review, we stated, “For us, hard pass on this. 8-12% for $2,200 is insane.”

And with an overclock back then, we were able to nearly equate the 3090’s performance with the OC 3080. That’s how close they were.

The odd 80 Ti/Super class from the 20 series to 40 series occupy the space between the 80 class and the flagships. There’ll likely be another between the 5080 and 5090. 

The Super refreshes really should be called the “oopsies” edition GPUs. NVIDIA rolls these out when they make an “oopsies” on price and public sentiment, using Supers to meet halfway on price. Our hope is that the 5080 and 5090 gap ends up again as an “oops, let’s fix this” rally from NVIDIA with a mix of the 2080 Super’s (watch our review) or 4080 Super’s relatively sane pricing along with the 3080 Ti’s (watch our review) aggressive configuration. That might start to help fix this a little bit. 

The 70 Ti/Super class drops hard. The 1070 Ti (watch our review) had a 68% CUDA core configuration for this class, falling to just 41% for the 5070 Ti (read our review). By this logic, the 1070 Ti offered far more GPU relative to the 1080 Ti than the 5070 Ti is to the 5090.

Next, we’ll expose NVIDIA’s grand switcheroo between the 70 and 80 class GPUs. 

From the 770 series to the 3070 (watch our review), the CUDA core count of the 70-class cards once reliably was between 53-59% of the flagship’s CUDA core count. 

Then the 4070 bore only 36% of the CUDA cores the 4090 had, and the card falling in the 50-60% range was now the 4080.

Moving into the present, the RTX 5070 has an anemic 28% of the flagship’s configuration. If you were to extend the 70-class line out on its previous trend, you’d arrive around the same position as where the 80-class is now. Strictly speaking in proportions and if we want to do funny percent math, the 3070’s (watch our review) core allocation relative to its respective flagship was 100% higher proportionality against the 5070’s.

The 60-class is where it gets really bad.

That 28% figure for the 5070 is lower than almost every 60 class configuration. The 60 class traditionally occupied the 30-40% range with a high outlier in the 20 series at 44%. This tracks with the fact that we were softly positive on the 2060 at its release – more positive than the 2080. The 3060 returned to the low 30% range, but the 4060 got slashed to 19%. Here’s what we said of the 4060’s worse cousin, the 4060 Ti, “The RTX 4060 Ti 8GB is one of the worst GPU launches from NVIDIA that we’ve ever covered.”

And that brings us to the 50-class. The 19% on the 4060 is where the 50-class has sat multiple times. NVIDIA covered this segment during the 20 series with the 16 series GPUs, which we didn’t plot for sake of simplicity. Moving forward, the 3050 was a 24% configuration, and it’s no wonder why the 4050 got canned – it would be like scraping the bottom of the barrel so hard that you just get splinters.

But what’s crazy is that the 5070 barely clears the 27% config of the old GTX 950. That’s just sad.

Now that we’ve established the trends, let’s keep all of that in mind and analyze pricing in the same way.

Chart - Inflation Adjusted Prices

This line plot tracks the launch price of all the same GPUs in each class, adjusted for inflation from the month of each GPU’s launch to January 2025.

Right away, we see that the flagship class has changed massively. The 780 Ti, 980 Ti (watch our revisit), and legendary 1080 Ti fall within a consistent $100 spread. The 980 Ti was slightly cheaper at a $650 launch price, which is $865 after the inflation adjustment. The 1080 Ti sits at $912, in stark contrast to the massive jump of the 2080 Ti at $1,510 adjusted. That’s a 66% cost increase gen-on-gen for the customer. It was technically available for $1,000, but in very limited quantities, and the vast majority went for $1,200, which is what we adjusted from.

The price went up again for the 3090, with a slight relief in the 4090, before jumping again to the $2,000 mark with the 5090. It undoubtedly costs more for NVIDIA to make a 5090 than it did to make a 1080 Ti, but there’s no argument that more than double the retail price is painful for a consumer.

The 80 class has also risen, though not to the same extreme degree as the flagship class. The GTX era 80s had inflation-adjusted prices between $734 and $886. There was a slight bump to just over $1,000 in the 20 series, followed by relief to the mid-$800s at the 3080, before rising insurmountably to the 4080.

When taken alongside the CUDA core configurations, all of this underscores both just how good the 3080 was and how terrible the 4080 was. The 3080 had a spike in core allocation and a return to “normal” pricing, while the 4080 fell off core config cliff and the price went up at the same time. Coming to the present, the 5080 is back at the same relative price as the 2080, but at a much worse relative CUDA core count.

The 80 Ti/Super class is an oddball – as if NVIDIA hasn’t been able to decide whether it’s better as a later, better value 80 class alternative like in the 20 and 40 series, or if it should be a weirdly positioned, poor value cash-grab like the 3080 Ti.

The 70 Ti/Super class has risen in price across the generations that it’s existed, from roughly $500 at its introduction in the 10 series to $849 in the 40 series. AMD Radeon GPUs were competitive in this price bracket back in the 10 and 20 series days, which is likely the reason why we see this aggressive pricing during that time period. From the 30 series onward, NVIDIA’s dominance has allowed this class of card, specifically, to sit comfortably between the 80 and 70 classes.

The 70 class has managed to stay relatively flat from one end of the chart to the other. The all time low price was in the GTX 900 series at $440, and the high point was the 20 series at about $750. That’s a large swing, but it’s stayed relatively flat since then.

The 60 class paints a similar picture. The inflation adjusted price line is generally flat overall with a slight downward trajectory since the 20 series, but in that same time the core config has gone into the dumpster. We don’t know anything about a theoretical future 5060, but we’d bet it won’t be a pleasant addition to this data set.

Finally, the 50 class hasn’t seen much action, but it hasn’t seen many releases in recent years – probably because the 4060 took its actual place. Judging by the 3050, NVIDIA is probably unwilling to launch a GPU for under $250 again, let alone the $145 mark of the 1050 (watch our review).

Additional Segmentation

Over the years, the means of product segmentation have migrated. Product segmentation isn’t inherently an evil thing, and especially in the world of silicon where the costs are enormous to make any of these products, but it can be applied in ways which just don’t feel good as a consumer. Segmenting the 1080 Ti at 11GB versus the Titan cards at 12GB didn’t feel particularly bad. It was obvious what they were doing, but the affected user base was much smaller.

Some of the other ways NVIDIA has historically segmented its products include splitting double precision out into only the highest-end cards, which at one point included Titans. Another is by forcing users over to Quadro for verified drivers as an additional layer of liability reduction for big organizations.

Neither of these two segmented features are noticeable to the vast majority of end users, so it doesn’t feel as bad to the consumer. Over time, that has drifted to VRAM increasingly, which now means there is a new developing class of users. 

For the gaming audience, we get situations where a $750 video card can find itself in situations of unplayable stuttering and latency nearing 800 ms PCL due to VRAM overload and swapping.

Joining the scientific user base that once needed double precision, or now might need various machine learning capabilities, there is now the segmented customer base of so-called “creators.” Not just YouTubers, but anyone making 3D art, games, or similar media.

These users are being pushed into the 90-class, which is further diminishing the capabilities of the highest-end gaming cards or pushing those high-end gaming consumers into price categories of professionals who use their GPUs to make money. It’s easier to shrug it off knowing it’ll make back the time, even if it’s still unpleasant.

“Arbitrary” Naming

Back in our RTX 4080 Problem video, we talked about how all of this is predicated on the assumption that the names mean anything. Like Whose Line Is It Anyway, sometimes it feels like the names are made up and the prices don’t matter.

We’ve been open about our opinions about this changing over the years: At one point, it did feel like names were somewhat arbitrary. It is just a name, and it’s ultimately the specs and price that matter. But the shift came over the last couple generations, where we came to appreciate that what’s in a name is important.

NVIDIA has used the 80-class cards to establish an expectation in customers, and regardless of whether NVIDIA intends it to still be perceived as the high-end as opposed to some mid-range card (which it is now), the fact is that their consumers do perceive the 50 name as intended to be high-end.

This is sort of a death of the author scenario, but then NVIDIA doesn’t want to name a $1,000 video card a “5070.” That creates new problems. 

NVIDIA has died as the author, and the consumer is now in control over what these names mean. To quote someone in the industry, it’s the “perception of reality” versus the reality. 

If NVIDIA wants to establish a reality where an 80-class card is half of a 90-class card, they can do that; however, if the end users perceive that a 5080 should be a true high-end device, that’s all that actually matters. NVIDIA is also responsible for this. The company spent a decade establishing the 80-class cards as the top-of-the-line, behind only the Ti class cards. It has now bifurcated those two lines and created a large gulf between them.

And this is getting worse with 5070 cards that are now more similar to older 50-class cards.

And so while the name itself is technically arbitrary as compared to the specs, the name matters. It defines an expectation. 

Let’s explore that philosophy a bit more. If Toyota suddenly starts shipping rebadged Yugos that it calls Camrys, that’s going to cause problems with the customer base. That’s what NVIDIA is doing. If the AMC Gremlin is sold under literally any name, it’s going to cause problems.

The point is, the RTX 5080 is a Yugo. Or a Gremlin. Or a Ford Pinto. And NVIDIA has spent a decade branding it as a supercar (and it was at one point a supercar). 

Conclusion

Zooming back out, we think the overall picture is clear. NVIDIA has downsized essentially all of its gaming GPUs in terms of relative configuration compared to each generation’s flagship. All of the lines go down. The chart from earlier had a lot of words to say one thing: Line go down = bad. We don’t want the line to go down. We want the line to stay the same or go up.

The 80 class is now in line with former 70 class GPUs and the 70 Ti/Super class is now in line with former 60 Ti class territory. The last 60 class card was configured like a 50-class of yore. 

Some might argue that the 4090 and 5090 being such monsters skews the comparisons, but we think that’s more of a perception issue based on NVIDIA's success at pushing the cost of the high-end higher. NVIDIA’s flagship GPUs have been very large pieces of silicon since the 20 series, and the CUDA config cutting had barely begun at that point, and the MSRP wasn’t as high as it is now.

The price of NVIDIA’s GPUs has generally gone up over time, even accounting for inflation that does things like turn the former $700 1080 Ti into a $912 GPU in today’s money. But then you look at $900 GPUs in today’s money and that’s a 9070 XT. And the 9070 XT isn’t positioned where the 1080 Ti was. The closest GPU might be the 5080 at $1,000 and that also doesn’t feel like a 1080 Ti by price. Flagships, however, are the worst, rising from that level to $2,000 with the 5090. Non-flagships haven’t risen quite as much, but it’s still significant. In this case, line go up = bad. For the consumer, anyway.

The 70-series here is one of the most textbook examples of shrinkflation. While the price point has stayed fairly consistent for a few generations, remember that the relative CUDA core configuration has dropped by a huge amount during that time. It’s gone from $610 with a 56% configuration in the 30 series, down to $550 with an embarrassing 28% core configuration in the 50 series.

NVIDIA is giving you a half-size slice of the GPU pie with the 5070 than it did with the 3070, but it’s charging you basically the same amount of money for the privilege.

All of the GPUs are victims of the configuration cutting we talked about. Even the technically-cheaper-than-they-used-to-be 70 and 60 class cards are providing less of a share of the capabilities of their respective flagships than they used to. 

And AMD isn’t immune to this, of course. We have an entire article dedicated to the company’s fake MSRPs that delves into this. NVIDIA, however, holds 90% of the market, and it’s important for you to understand how your money is disproportionately losing value when it’s spent with NVIDIA versus many years ago.

We don’t have an answer for this. It’s sort of too big, but it’s important to know about and to start thinking about. Maybe enough people will pay attention to this so that it will help them make informed purchasing decisions.