Nvidia partners are already preparing for the GeForce RTX 60 Series even though the 50 Series has yet to launch officially. These trademark registrations show the models some brands are expecting next.
According to an EEC (Eurasian Economic Commission) filing from Sinotex International Industrial Ltd, the brand expects Nvidia to launch more than a fair few GPUs. These include models from the upcoming RTX 50 Series, such as the rumoured RTX 5090, plus others from the following RTX 60 Series. Some variants we’ve not yet seen among the RTX 40 lineup, like the RTX 4050 and RTX 4090 Super.
Since these are just expectations via trademark registrations, there’s nothing to say Nvidia is working to expand current-gen Ada Lovelace entries. You may as well consider them a rumour.
Talking of which, recent reports claimed that RTX 5090 could offer double the performance of RTX 5080, or a 32% uplift over RTX 4090 when factoring in the CUDA core difference alone. But that will come at a steep price – above $2,000, which is higher than many complete gaming systems. That said, I bet RTX 5090 will nonetheless fly off shelves if it offers such a generational boost.
With RTX 60 Series being so far into the future, we’ve not heard much about it aside from its potential use of Nvidia’s Rubin architecture. That, in itself, remains a mystery in 2024. Will they also offer a substantial generational uplift over Blackwell? Your guess is as good as ours.
If Sinotex does have an inside track, these trademarks could indicate Nvidia’s strategy to dominate the high-end with many Super and Ti models. Supposedly, Team Green won’t face much competition in this segment, as AMD’s upcoming Radeon 8000 Series and Intel Arc Battlemage GPUs will instead shift focus to entry-level and mid-range options. All we know for sure is that this indicates partners are preparing for RTX 50 Series in some shape or form.
It’s a little late in the day for another RTX 40 Series card with the imminent release of its successors, but new entries into previous-gen ranges are all the rage right now. AMD has seen some success by revisiting its Ryzen 5000 series and introducing 3D V-Cache on top, and Nvidia itself launched a 6GB RTX 3050 earlier this year.
Given track records, RTX 4090 Super seems much more plausible, particularly factoring in the performance that’s left on the table. The original RTX 4090 uses a cut-down AD102 GPU packing 16,384 CUDA cores. In contrast, the full AD102 chip houses 18,432 CUDA cores, leaving room for a faster card. The latter could be around 12% quicker if we base our calculations on the CUDA count alone.
In any case, we should get more information during CES 2025, so stay tuned.