Although still yet to be announced, Nvidia’s RTX 5060 Ti is coming — and according to leakers, it’s coming soon. The card is said to be launching in two different models, much like the RTX 4060 Ti. However, the RTX 40-series equivalent really struggled to find its place among the best graphics cards.
I’ve been keeping tabs on various leaks related to the RTX 5060 Ti, and although there are some good news, it does seem that Nvidia is essentially creating an updated carbon copy of the RTX 4060 Ti. If there’s one party that will benefit from such a decision, it’s AMD with its competing graphics cards. For gamers, the RTX 5060 Ti might end up being a GPU that sounds great on paper but comes with its own challenges.
The RTX 5060 Ti will have a lot in common with its predecessor
Nvidia’s RTX 50-series is only comprised of four GPUs right now: The RTX 5090, the RTX 5080, the RTX 5070 Ti, and the RTX 5070. The last one isn’t out yet, but should be out in a matter of days, launching on March 5. The former three are already here, although plagued by a host of problems, ranging from missing ROPs to awful availability and scalper pricing.
Get your weekly teardown of the tech behind PC gaming
Issues aside, it’s clear that Nvidia will still expand the Blackwell generation with at least two more desktop GPUs. The RTX 5060 Ti and the RTX 5060 are said to be launching in late March and April, and Wccftech has now revealed a little more about the specs for the more expensive RTX 5060 Ti.
The RTX 5060 Ti is said to run on the GB206 GPU and will use the PG152 PCB. Wccftech confirms what many of us have long been expecting: Nvidia is readying two variants, one referred to as “SKU 10” and the other dubbed “SKU 15.” The difference lies in the amount of video memory for each card; SKU 10 refers to the 16GB model, and SKU 15 means the 8GB VRAM variant.
Now, we’ve got good news and bad news. The good news is that Nvidia is giving the RTX 5060 Ti the same GDDR7 bandwidth boost that the rest of the Blackwell lineup is already using. This means switching to 28Gbps GDDR7 VRAM modules, and as a result, up to 448GB/s of total bandwidth, marking a staggering 55% uplift over the RTX 4060 Ti.
On the other hand, the memory bus remains at 128-bit for both the 16GB and the 8GB models. Using such a narrow memory interface is almost guaranteed to affect performance, and even Nvidia’s 16GB of VRAM won’t save that. We’ve seen this time and time again in GPUs such as the aforementioned RTX 4060 Ti or AMD’s RX 7600 XT.
Moving on, both cards are said to come with a total board power (TBP) of 180 watts, which is a 20 to 15-watt increase over the previous generation. Wccftech also claims that Nvidia and its partners will use both the 8-pin and the 12V-2×6-pin connector, varying between different models.
As for the rest of the specs, the RTX 5060 Ti is rumored to come with 4,608 CUDA cores and a maximum clock speed of 2,520MHz. These specs haven’t been floating around nearly as much as the rest of this, though, so make sure to take it all with a bit of skepticism.
Assuming all of the above turns out to be true, the RTX 5060 Ti will share a lot of specs with its predecessor. The RTX 4060 Ti offered the same amounts of VRAM across the same, narrow 128-bit bus; it was also very close in terms of CUDA cores, sitting at 4,352 CUDAs, and had a nearly identical clock speed.
This isn’t just a recipe for minor generational upgrades. It’s a mistake that, once made, Nvidia may now be choosing to repeat.
The VRAM problem
It’s no secret that the RTX 4060 Ti had a VRAM problem. Depending on the model, the issue was twofold: It either didn’t have enough VRAM, or it didn’t have a robust enough memory interface to support the 16GB that it was given. In both cases, this setup resulted in underwhelming performance.
We’ve benchmarked the RTX 4060 Ti and its many rivals ourselves, and the results were similar across the board: the lack of VRAM stifled an otherwise solid GPU in a major way.
Take, for instance, our comparison between the RTX 4060 Ti and the RX 6700 XT. One would think that a match-up between those two GPUs would end in a win for the RTX 4060 Ti — after all, it belongs to a newer generation of graphics cards. However, AMD beat the RTX 4060 Ti in most titles across our comprehensive test suite, scoring huge wins in titles such as Forza Horizon 5 and The Last of Us Part 1.
On average, the 8GB RTX 4060 Ti was slightly faster than the RX 6700 XT, but there are titles where the previous-gen AMD card tied with the Nvidia GPU. In some, it won, including that 32% lead in The Last of Us.
That success can largely be traced back to VRAM. The RX 6700 XT comes with 12GB of memory — a notable increase over the 8GB version of the RTX 4060 Ti. It also has a 192-bit memory bus, allowing for much better bandwidth. As a result, games that benefit from more video memory immediately breathe a sigh of relief when played on a GPU that can support those higher VRAM needs.
Of course, things looked worse when ray tracing was a factor. Nvidia beats AMD in that regard, no contest. But for gamers who just want stable gameplay, it was interesting to see an older card beat the newer RTX 4060 Ti simply on the basis of more VRAM.
The RTX 4060 Ti also became famous for the fact that it was dangerously close to its predecessor in terms of performance. In VRAM-constrained titles, such as Hogwarts Legacy, we found the two GPUs head-to-head, with the RTX 3060 Ti winning by a small margin. Although both cards shared the same 8GB of VRAM, the RTX 3060 Ti had a lot more bandwidth to work with thanks to the 256-bit memory bus.
As more and more games call for more than 8GB of VRAM (see Indiana Jones and the Great Circle for a recent example), the RTX 5060 Ti with 16GB of memory would be a fantastic option. The only downside is that Nvidia is said to be using the same 128-bit bus, and that will result in those same memory constraints once again rearing their ugly heads.
The VRAM problem may live on in Blackwell, and that’s despite the AAA gaming world moving forward at a rapid pace.
Why is this good news for AMD?
If proven true, Nvidia’s resolve to stick to that less-than-perfect memory setup may spell good news for AMD. Not just AMD, but even Intel.
No GPU exists in a vacuum. The RTX 5060 Ti, when it launches, will face scrutiny and be compared to the likes of the RX 9070 or the RX 7800 XT. Both GPUs have more VRAM, so I wouldn’t be surprised if that RTX 4060 Ti versus RX 6700 XT comparison will make a comeback in this generation, too, but this time with newer versions of those cards.
AMD has another ace up its sleeve: The RX 9060 XT. It’s currently unclear when that card might launch or what kind of specs it might have, but if AMD can provide more memory or better value, it’ll be a contender against the RTX 5060 — and that’s not too far off from the RTX 5060 Ti. It should also be significantly cheaper. The RTX 5060 Ti will likely be in the $400 to $450 price range; AMD’s counterpart could be as cheap as $300.
Historically, AMD cards tend to offer more VRAM at more reasonable prices. Take the RX 7800 XT or the RX 7900 GRE, for example, but even the RX 7700 XT: All of those GPUs have 12GB VRAM or more. By comparison, Nvidia keeps its most mainstream-oriented offerings starved for video memory.
Meanwhile, Intel is not much of a contender in the GPU race, but it shares similar values to AMD: It’s laser-focused on performance-per-dollar. Its GPUs, while budget-friendly, offer more memory, which helps them face newer titles.
In all honesty, the RTX 5060 Ti might turn out to be just fine. Sure, the 8GB VRAM version may suffer from not having enough memory for AAA blockbusters; meanwhile, the 16GB VRAM variant will still have a far-too-narrow memory bus. But switching from GDDR6 to GDDR7 will still provide a huge uplift in bandwidth, and that will make up for that 128-bit bus — to some extent. Still, it’s hard not to imagine the performance of a 16GB, 192- or 256-bit, 28Gbps RTX 5060 Ti. That’d be one beast of a mainstream GPU.
The next couple of weeks will give us a better hint of what to expect. I don’t mean just my expectations for the RTX 5060 Ti, because I’m keeping those really low. I mean the general state of the GPU market, including GPU availability, pricing, and performance.
With the RTX 5070 and the RX 9070 XT soon set to break cover, we’ll have a better idea of where the RTX 5060 Ti might land in the grand scheme of things. There’s one thing I’m afraid of, though: It might not offer the impressive generational leap many gamers are hoping for.
Read the full article here
