Why you can trust Tom's Hardware Our expert reviewers spend hours testing and comparing products and services so you can choose the best for you. Find out more about how we test.
Nvidia GeForce RTX 5060 Ti 16GB: Great at MSRP, but retail prices could be much higher
Nvidia's RTX 5060 Ti 16GB delivers a solid combination of performance for the suggested $429 base MSRP. However, as we've seen with every other GPU launch of the past five months, retail prices can be much higher. It's impossible to separate performance from pricing when looking at the overall value of a GPU, and the only thing concrete that we can point to are the MSRPs. Except those can run the gamut from being at least moderately accurate to being completely nonsense.
Where will the RTX 5060 Ti 16GB actually land? We suspect it will trend closer to $500 in the near term, but supply will have a lot to say about that. We've seen the RTX 5070 sell for $550–$600 over the past few weeks, and it's substantially faster than the 5060 Ti 16GB — about 30% faster overall. If we put $600 down as the nominal price for the 5070, then the 5060 Ti 16GB needs to stay well below $500 to offer a similar value. But both GPUs could end up at much higher prices, so you'll just need to see what's available when you're ready to buy.
When the 4060 Ti 16GB came out a month after the 8GB variant, it felt severely underwhelming. Neither version was really designed to handle 4K gaming, but that was the only place where we measured a significant difference in performance. Two years later, things haven't changed too much, but the reduced $50 price gap (on paper at least) between the 5060 Ti 8GB and 16GB makes the 16GB a far easier recommendation. In fact, we'll go so far as to question why Nvidia even felt the need to create an 8GB version.
Yes, 8GB will be cheaper, and it will also be more limited due to the lack of VRAM. There are games (Indiana Jones and the Great Circle) where you can't even try to run ultra settings on an 8GB card. That's an Nvidia promoted game that simply crashes to desktop with a video memory error when you try higher settings on the 4060 and 4060 Ti 8GB GPUs, along with a bunch of other previous generation RTX cards.
The real market for the 5060 Ti 8GB likely isn't the DIY or gamer crowds. Instead, it will probably find its way into prebuilt PCs, where the big OEMs want the highest model number possible at the lowest price. Shaving $25 off the bill of materials for a $1,200 desktop by using an 8GB 5060 Ti instead of the 16GB model? We'll probably see a lot of that. We'll be looking to acquire and test an RTX 5060 Ti 8GB as soon as possible to show exactly how that will affect gaming in the near future.
The good news with the 16GB card is that memory bandwidth has improved thanks to GDDR7, so that it's not likely to hit VRAM capacity or bandwidth limitations. 56% more bandwidth than the 4060 Ti is a sizeable improvement. The fact that most games only show about 15% higher performance indicates that GPU compute is the limiting factor more than bandwidth, however.
We also wish Nvidia has simply opted for an in-between solution. It's using 24Gb (3GB) GDDR7 chips on the mobile RTX 5090, as well as the RTX Pro 6000 Blackwell server, workstation, and data center variants. Obviously those chips exist and appear to work just fine. How much would it have cost to use four 3GB chips on one side of the PCB, rather than putting eight 2GB chips on both sides of the PCB in clamshell mode? We can't help but think the 16GB configuration costs more, so perhaps there just isn't enough 3GB chip supply right now to cover all the products that would benefit.
And so we end up with the weird bifurcation where you have either too little VRAM in the case of the 8GB model, or more VRAM that the GPU really needs in the case of the 16GB card. But we'd rather have too much than not enough. If you're interested in buying the RTX 5060 Ti, we would strongly advise potential customers to get the 16GB model this round. There are simply too many 'edge' cases where 8GB isn't enough, and they're becoming increasingly common. $50 extra is money well spent.
But as we've already said numerous times, the price difference could very easily end up being more than $50. And factors like on again/off again tariffs, limited supply, product demand, and more could push the 16GB card to the point where maybe it won't be the better choice. The RTX 5070 still serves as a ceiling on how much more the 5060 Ti 16GB can realistically cost before it's "too much," but with 5070 cards often listed for $700 or more, there's a lot of wiggle room right now.
Price and availability will be the key determiners of how good the 5060 Ti 16GB looks, and that will also vary by market. Europe and Asia might end up with a much different GPU landscape than the U.S. as far as graphics card values go.
What we can say is that the 5060 Ti 16GB isn't a massive generational improvement, but it is an improvement. It's also supposed to be less expensive than its 4060 Ti 16GB predecessor. Those are both good things, and stuff like neural rendering, DLSS 4, and Multi Frame Generation are merely extras that you can use as you see fit. Now we just wait to see what today's launch looks like, how quickly the 5060 Ti models sell out, and how high prices go.
Our score of 4-stars represents a "best guess" on what the 5060 Ti 16GB will look like in the current GPU market. Obviously, prices for all graphics cards, new and used, are all over the map. If the RTX 5060 Ti 16GB costs 50% more than the MSRP, and other cards don't show a similar markup, that makes it a worse value and a less desirable card and it would deserve a lower score. We can't predict where things will go, so pay more attention to the performance and real-world pricing than the single score that we've assigned, because uncontrollable factors play into the overall package.
- MORE: Best Graphics Cards
- MORE: GPU Benchmarks and Hierarchy
- MORE: All Graphics Content
