Remember, that Super teaser that NVIDIA released right before Computex? Yeah, the one with the shiny silver plate with “Super” engraved on it. So, turns out whatever that something super is hasn’t been revealed yet, and NVIDIA is prepping two new Turing cards, faster than the GeForce RTX 2070 and RTX 2080 for E3, right in time to crash AMD’s Navi GPU launch.
This isn’t surprising at all, to be honest, NVIDIA has had a history of countering disruptive Radeon launches with unexpected “Ti” cards. The prime examples are the GTX 780 Ti, the 1070 Ti and more recently the GTX 1660 Ti. These cards weren’t based on new architectures or anything, they just had more shaders and/or memory than their non-Ti counterparts.
At their Computex keynote, AMD showed off the new RDNA GPU architecture, designed exclusively for gaming. The same macro-architecture is going to power the PS5 and the Xbox Scarlett consoles, as well as the upcoming Navi cards, AKA, the Radeon RX 5700 series. At AMD’s press conference, the company claimed that the higher-end Navi will be up to 10% faster than the RTX 2070. However, considering that only one benchmark was shown says a lot. The most likely scenario is that the RX 5700 (like the Radeon VII goes up against the RTX 2080) will compete with the RTX 2070 possibly (hopefully) at a lower price.
NVIDIA has held a very commanding position in the GPU market for years now, and I’m sure it’d like to keep that as is. As mentioned in an earlier post, the GDDR6 memory on the RTX cards is clocked very conservatively. There’s a very good chance that these new Turing cards are faster variants of the RTX 2070 and 2080, with faster 16Gbps VRAM chips and potentially higher core clocks.