When the 4060 Ti launched, it wasn’t exactly well-received, and for multiple reasons. The first being that the 8GB model was released to start, sporting a 128-bit bus for $399, and led to a lot of criticism from reviewers and end-consumers, even before it launched. Two months later, NVIDIA tried to rectify the issue with the 16GB variant, but it featured the same memory interface and a price premium of $100 more, putting it closer to RTX 4070 territory. So a generation on, we now have the launch of the RTX 5060 Ti 8GB and 16GB models, but for slightly more attractive price points, but with the rest of the 50 series focussing heavily on AI upscaling, are we going to see an improvement over the predecessors for a lower price point, or are we just going to see more of what the rest of the Blackwell lineup has given us? Including inflated pricing that throws MSRP’s swiftly out of the window? Well, that’s what we’re going to find out.
So, the main points I’ve kind of touched on, and the big sticking point really came down to memory on the Ada Lovelace based predecessors, but this is the opportunity for NVIDIA to mix things up, and get some good press on what is arguably going to be their most important cards in the lineup, as the pricing is hitting that sweetspot that a lot of gamers see as viable options. With the RTX 5060 Ti 8GB and 16GB, we’re looking at pricing of $379 and $429, which means we’re seeing a $20 cut from the 8GB 4060 Ti, and a more significant $70 drop from the 16GB 4060 Ti, and this also means that gamers are only expected to pay $50 more for double the memory, which in all honesty, is much better than what we saw before, and much more palatable.
Sadly, though, if the rest of the 50 series lineup has taught us anything, then we know those prices are for the most part, a bit of a pipedream. Don’t get me wrong. I’m not denying that they will be available, but likely only a handful at select retailers, before we see pricing jump up beyond $100 more than they should be, and even at $100 more, that’s me being somewhat generous, as we’ve typically seen other Blackwell cards commanding a lot more over their recommended retail price.
Pricing aside, being a Blackwell based GPU, we do get all of the latest tech, from the latest RT cores, 5th generation Tensor cores, and DLSS 4 functionality, which along with it, brings multi-frame generation and a new Transformer model to not only increase performance in conjunction with Reflex to reduce latency, but also improve quality at the same time. One of the big things for me though is the specs, because the most important one is that this card, and its 8GB sibling, come with GDDR7, which offers a large uptake in bandwidth compared to what we’ve seen previously but sadly we still see a paltry 128-bit memory interface, though I can only assume that with the inclusion of GDDR7, NVIDIA are maybe hoping that the lower bus won’t be as much of a detriment to performance, like we saw on the 4060 Ti where gamers screamed about having a 16GB model, only to be let down due to throughput.
Obviously, Blackwell works a lot differently, and with improved cache processing, this shouldn’t be as bad an issue, but before any release, numbers are the thing that really gets people’s attention, and 128-bit will be getting it, but likely for the wrong reasons.
