The RTX 5060 Ti is the next in Nvidia’s line of Blackwell GPUs, with the company bringing two variants to the table today: an 8GB version and a 16GB version. Targeting 1440p, the RTX 5060 Ti brings Blackwell and its AI-powered advances to the sub-$500 MSRP price point, but how does it stack up overall? Here is our review.
Specs:
- Architecture: Blackwell
- CUDA Cores: 4608
- Tensor Cores (AI): 5th Generation 759 AI TOPS
- Ray Tracing Cores: 4th Generation 72 TFLOPS
- Boost Clock: 2.69 GHz (reviewed)
- Base Clock: 2.41 GHz
- Memory: 16 GB/ 8 GB GDDR7
- Memory Bandwidth: 448GB/s
- Memory Interface: 128-bit
- I/O: 3x DisplayPort 2.1b; 1x HDMI 2.1b
- Max GPU Temp: 87 degrees Celsius
- Total Graphics Power: 180 Watts
- Required PSU: 600 Watt
- Power Connector: 1x PCIe 8-pin (reviewed PNY version)
- MSRP: $429 (16 GB reviewed); $379 for 8 GB variant
PNY RTX 5060 Ti OC 16GB First Look
The RTX 5060 Ti doesn’t come in a Founders Edition board, so our review this time around comes in the form of an AIB model from PNY. The PNY RTX 5060 Ti OC 16 GB. Despite being an overclocked variant of the RTX 5060 Ti, this 16GB board is set to sell at the MSRP of $429, though during a press briefing Nvidia said that this price doesn’t include any impact the current Presidential administration’s tariff plans could have on the price here in the US.
The PNY RTX 5060 Ti OC 16 GB is a two-slot card, and even by 50-series standards it's rather small. Sporting a dual-fan design to keep the card cool under pressure, the GPU feels perfect for small form-factor builds thanks to its compact design. Crucially for many people looking to upgrade from a previous generation of GPU, the RTX 5060 Ti comes with your standard 8-pin power connector instead of the 12VHPWR connectors the rest of the 50-series has, making upgrading much easier (and likely giving some piece of mind to those concerned about reports of melting cables the last two generations).
The RTX 5060 Ti will benefit from the bevy of technologies brought to bear by the Blackwell architecture, namely the advancements with Nvidia’s image reconstruction tech, Deep Learning Super Sampling, or DLSS. This includes features such as Multi-frame generation, as well as future features such as Neural Rendering in the form of RTX Skin, Neural Shaders, RTX Mega Geometry, and much more.
We did a larger breakdown of Blackwell following its reveal back at CES 2025 in January, which you can read here.
The obvious elephant in the room is that Nvidia are launching two models of the RTX 5060 Ti, an 8 GB version and a 16 GB variant. We’ve got the 16 GB version for our review today, and Nvidia did note during a press briefing ahead of this review embargo that the 16 GB variant would be more prevalent in the wild. Admittedly, the 16 GB version is the more compelling offer on paper, despite costing $50 more than the 8 GB (at least at MSRP). If you’re in the market for a GPU in this price range, it’s something to consider as the vast majority, and maybe even possibly all the reviews you read today will be of the 16 GB version.
That said, if you’re looking at doing any heavy ray tracing applications, the extra VRAM will be needed, and as last generation’s RTX 4060 Ti 8 GB taught us, that amount of VRAM isn’t enough even when leveraging DLSS in titles. The 8 GB version has an MSRP that is actually lower than the retail price of last generation’s RTX 4060 Ti 8 GB version ($379 vs $399, respectively), but that is also assuming you can nab any of these cards for their retail prices in today’s economy.
There is also the worry of future proofing - 8 GB just might not be enough for the next generation slate of games we’ll be seeing in the years to come. But it’s an option, and having more options is never inherently a bad thing.
The PNY RTX 5060 Ti 16 GB OC version is an overclocked variant of the standard 5060 Ti spec. This includes a boost clock OC up to 2.69GHz on the PNY versus the base spec’s 2.57GHz boost clock. This extra boost is nice to see on a card that, in a perfect world, is priced at Nvidia’s MSRP for the 16 GB card.
PNY RTX 5060 Ti 16 GB OC Synthetic And Gaming Benchmarks
Putting the GPU to the test, we ran the RTX 5060 Ti through our regular suite of gaming and synthetic benchmarks, focusing squarely on 1440p for our testing. We tested both traditional raster and ray tracing tests, alongside 3DMark’s suite of synthetic benchmarks.
We’re using our Starforge Systems Voyager II test bench as we have all year for our GPU reviews. We detailed why in a separate post, but here are the specs for a quick glance.
- CPU: Intel Core i7-14700K
- CPU Cooler: Starforge Custom Bitspower 360mm Liquid Cooler
- RAM: Teamgroup Delta RGB 32GB DRR5 6000 CL38 (2x16GB)
- Motherboard: MSI Z790 Tomahawk Wifi DDR5
- Cables: CableMod Pro ModMesh Sleeved Cable Extensions (Black)
- Primary Storage: 2TB PCIe 4.0 NVMe SSD (min 7000 read)
- Power Supply: Corsair HX1200i 1200 Watt Platinum
- RGB Fans: 6 x Bitspower 120mm ARGB
- Case: Lian Li O11D EVO RGB Mid-Tower (SMITE Custom Printing Version)
Being a 1440p-focused card, we put the RTX 5060 Ti up against other similarly-targeted cards across Nvidia and AMD’s line up of GPUs.
Nvidia:
- ASUS Prime RTX 5070 Ti
- RTX 5070 Founders Edition
- RTX 4070 Super Founders Edition
- RTX 4060 Ti 8GB Founders Edition
AMD
As with all our recent GPU reviews, we want to test these cards as people will actually use them. With this in mind, when testing ray tracing loads, we’re leveraging both DLSS and AMD’s solution, FidelityFX Super Resolution when possible. Additionally, frame generation is enabled when possible because, again, this is how the majority of players are going to be actually using these cards. Like our previous 50-series reviews, we spun off DLSS 4’s Multi-frame generation testing separately, comparing the RTX 5060 Ti to its more powerful cousins.
When we could we utilized in-game benchmark tools in order to get the most consistent result possible in our testing. When we couldn’t we used a predefined, repeatable circuit to try to ensure consistency and accuracy when we could, such as a romp through the Defense Against Dark Arts tower in Hogwarts Legacy or measuring the same perspective throughout a replay of a Marvel Rivals match. ReSizeable BAR and XMP profiles were both enabled via the system’s BIOS.
PNY RTX 5060 Ti OC 16 GB Synthetic Testing
In synthetic testing, the RTX 5060 Ti lags behind every card in our bench aside from the last generation RTX 4060 Ti 8GB. In Firestrike, the RTXi 5060 Ti is 34% better with its Overall score compared to the RTX 4060 Ti 8GB, though it does lag behind last generation’s 1440p performer from AMD by 19%, the RX 7800 XT.
TimeSpy shakes out similarly. The RTX 5060 Ti performs better than the RTX 4060 Ti, though struggles against everything else in our bench, with the RTX 5070 FE besting it by 26% while the RX 7800 XT comes out 14% better, comparatively.
Port Royal, the ray tracing benchmark, shows Nvidia’s edge in its testing, as it’s the only synthetic test where the RTX 5060 Ti beats the RX 7800 XT, though by only 1% so it’s well within a margin of error.
PNY RTX 5060 Ti OC 16 GB Rasterized Benchmarks
When looking at rasterized gaming tests, the pattern holds up across the board. The RTX 5060 Ti puts up wins compared to the RTX 4060 Ti, such as 42% increase in performance at 1440p in Cyberpunk 2077 and a 41% increase in Black Myth Wukong. On average, at 1440p, our PNY RTX 5060 Ti OC 16 GB card performed 25% better across the board compared to the RTX 4060 Ti 8 GB we had on hand to compare with.
When put up against the AMD RDNA 3 1440p hitter, the RX 7800 XT, things don’t fare as well for Nvidia’s new GPU. Equipped as well with 16GB of VRAM, the RX 7800 XT performs on average 11% better than the RTX 5060 Ti in our testing,and that gulf could have been widened if Black Myth Wukong didn’t hate AMD silicon. Games like Final Fantasy XIV: Dawntrail see the RX 7800 XT leading the RTX 5060 Ti by 18%, and Cyberpunk 2077 ekes out an AMD win here, with the RX 7800 XT 12% faster over the RTX 5060 Ti.
Compared to the other 50-series GPUs in our bench, the more expensive RTX 5070 leads the RTX 5060 Ti by about 20% despite costing 27% more at MSRP (at least on paper). Meanwhile, AMD’s new RDNA 4 RX 9070 XT and non-XT cards continue this trend in our test, besting the RTX 5060 Ti by 31% and 28% on average, respectively.
PNY RTX 5060 Ti OC 16 GB Ray Tracing Benchmarks
When flipping on ray tracing and the software goodies that come with it, Nvidia’s RTX 5060 Ti does start to fare better compared to its competition. The 4th Generation RT cores start to show their worth compared to the RTX 4060 Ti and the RX 7800 XT, and they do help to close the gap compared to the RNDA 4 cards from AMD.
On average, the RTX 5060 Ti OC 16 GB card is 35% faster than the RTX 4060 Ti 8 GB (without the use of MFG), and it jumps to a 34% faster than the RX 7800 XT in our testing. However, compared to the RTX 4070 Super FE, the RTX 5060 Ti performs about 14% slower than the last-gen card. Typically, consumers can expect to see the lower-badge card on the newest generation of GPUs match or best the performance of the GPU one badge above it from last generation, but that doesn’t seem to be the case here.
PNY RTX 5060 Ti OC 16 GB DLSS 4 Tests
When looking at how the new GPU handles Multi-frame generation, we put it through the same suite of DLSS 4 tests we’ve done since the RTX 5090 Founder’s Edition review back in January. We opted not to include Alan Wake 2 and Star Wars Outlaws this time around simply due to time constraints, but we have included Cyberpunk 2077, Marvel Rivals, Hogwarts Legacy and Dragon Age: The Veilguard in our testing.
It’s here where you can start to see the RTX 5060 Ti able to punch a little bit above its weight class, though the experience isn’t necessarily the greatest across the board. Multi-frame generation can give some great framerate results, but that doesn’t always mean the input latency will be worth the hassle, as is the case in Cyberpunk 2077. Despite being able to achieve 89 FPS at 4K using full path tracing when using DLSS 4’s Multi-frame generation, the input latency skyrocketed to over 100ms. This is because the input framerate is so low at this resolution, MFG simply cannot make up for it.
Things were much better at 1440, with 160fps as our average with a 1% low of 87, which halved the input latency to about 54ms - which was still playable, especially in a single player game and one you’re likely to kick back and use a controller with anyways.
Marvel Rivals is the true test of input latency and how it is affected by Multi-frame generation, and I have to say the RTX 5060 Ti did handle it well. With an average of 156fps with MFG 4x turned on, it handled quite smoothly, and sure, the latency might be more than a competitive player would like, but if you’re a casual scrub like myself it was perfectly fine.
Dragon Age: The Veilguard had the highest 1% lows in our DLSS 4 tests, which made for the smoothest experience overall as I romped through the camp I’ve come to use as my testing ground.
In order to keep the comparisons as close as we could to the roster of 50-series GPUs this year, we opted to test at 4K, especially since MFG should be the special sauce to allow for this card to truly punch above its weight class. These are worst-case scenarios pushing the card to it’s actual limit. While it’s possible to play at this resolution when using DLSS 4 Performance with MFG set to 4x, that input latency will go down if the ingest framerate is much better, so I would recommend sticking with 1440p here and adjusting the MFG level as you need to fit your use case.
PNY RTX 5060 Ti OC 16 GB Thermals And Power Consumption
The RTX 5060 Ti is rated for 180W, and during our testing using PCAT, we never really broke that barrier. We did come close, with Cyberpunk 2077 drawing 179 watts of power from the PSU during our DLSS4 testing, but it’s good to see that even with the factory overclock from PNY, it doesn’t draw more power than it should. This is also the same test where we saw our highest clock speed, with the PNY RTX 5060 Ti OC 16 GB hitting a speed of 2.8GHz, much higher than the rated 2.69GHz on the box.
Thermally, the metal backplate and flow-through design of the board helped to keep temperatures cool. The RTX 5060 Ti hit a max temp of 70 degrees Celsius during our testing in Dragon Age: The Veilguard but otherwise throughout the use of the card it hovered around the 67 degrees Celsius mark on average.
PNY RTX 5060 Ti OC 16 GB Conclusion
So what do we make of all this? On average, the only card around its listed MSRP the RTX 5060 Ti beat on average was the RTX 4060 Ti 8 GB Founders Edition. And to be frank, this was expected, especially thanks to the larger pool of RAM the newer card had at its disposal. We didn’t have an RTX 4060 Ti 16 GB to test, which would have been a bit more compelling, but we still see a respectable 25% jump in performance at raster and 35% better performance when ray tracing on the new card compared to the older.
This stands to reason that this could scale the farther back you go, which is something Nvidia is keen to point out as so many players are using 60-class cards across the older generations of Nvidia GPUs.
While we don’t have a GTX 1060 to test Nvidia’s 50x performance claim it made in its marketing leading up to today’s launch, I will say with confidence that the RTX 5060 Ti will give you just night and day better performance over that older card - and that’s even before you turn on DLSS, something the 10-series doesn’t have access to.
However, it’s harder to justify the value (assuming this card is available at retail price) when compared to the other cards closer to it in performance in this stack. At $429, it has some competition with its last-generation counterparts, even if you’re looking at a used 7000-series AMD or a 40-series Nvidia GPU.
This is where reading other reviews to get a broader picture is going to be helpful, especially since we didn’t have a 16 GB 4060 Ti to compare against, and the time constraints to hit the embargo date put on this review limited the amount we were actually able to test and retest ahead of publication.
The RX 7800 XT, a card AMD marketed last generation as a 1440p outperformed the RTX 5060 Ti on average by 11% in raster. Nvidia’s strong lead in ray tracing is the saving grace for the RTX 5060 Ti, as it was 34% faster on average than the AMD card. Compared to the RTX 4070 Super FE, the RTX 5060 Ti 16 GB was 15% slower in raster and 14% slower in ray tracing, handily being beaten by an older card with less (and slower) VRAM.
The 60-class cards have always been propped up as the budget class, as well. And very early on in the Maxwell and Turing days that was certainly the case. But it’s harder to make the argument to someone who is upgrading from a card they might have spent $200 on years ago to buy a $429 - again, assuming perfect market conditions. At this price point, buying a console is as compelling, if not more so compelling an option.
When you can buy a whole console for the price of one component, it makes it hard to recommend, especially for gamers looking for the budget option here. If you’ve got a thousand dollar budget to spend on an RTX 5080 and so on, that argument makes less sense - you are locked into PC and are buying that more expensive card for a reason.
But if you’re someone who has held onto an old GTX 1060 or RTX 2060 for years, the RTX 5060 Ti is a good upgrade, but so is a PlayStation 5, or even potentially the upcoming Nintendo Switch 2, which is priced at only $20 more.
Then we get to the current situation the market is in where no GPU this generation has really been available at its retail price. Retailers have struggled to keep them in adequate stocks for people to buy them (indeed, at the time of this writing, Newegg is out of stock of many of its MSRP or close to MSRP models, as an example), and we have no clue what the current trade war climate is going to do to both pricing and availability moving forward.
That’s not Nvidia’s fault, mind you, but it certainly is a factor that consumers need to consider.
At the end of the day, the PNY RTX 5060 Ti OC 16 GB isn’t a bad card, per se. It does deliver on more performance over the 60-class card from the previous generation, and it provides a great playable experience at max settings when playing at 1440p, exactly what Nvidia set out to do here. In the right situation, Blackwell features such as DLSS 4’s MFG can add to the experience, giving higher framerates in some of the most visually demanding titles like Cyberpunk 2077. And being able to leverage the Transformer model of DLSS makes for an overall cleaner, more pixel-perfect reconstructed image that only benefits users.
If you’re determined to stay on PC and have been holding onto an older GPU, assuming the RTX 5060 Ti is in your budget, it’s a good upgrade. However, it’s hard to recommend if you cannot find it at its retail value - and even harder to determine its overall value proposition without the less expensive 8 GB version to test, or the upcoming RTX 5060 8GB card set to release next month.
Full Disclosure: The product reviewed was provided by the manufaturer for the purposes of this review.
Editor's Note: An original version of this review listed it as 6.5 when it was supposed to be listed at 7. We have corrected this and apologize for the error/confusion.
Beta Realms are Offline - 03:30 CEST
Beta Realms are Offline - 6:30 p.m. PDT
