Along with the first models of the new RTX 40 series, Nvidia has also unveiled the next generation RTX 6000 graphics card, which is specifically aimed at creative professionals. The RTX 6000 with "Ada Lovelace" architecture is the successor of the RTX A6000 with "Ampere" architecture (only the dropped "A" distinguishes the two model names) and has 48 GB ECC VRAM like the latter. However, the new RTX 6000 not only surpasses the GeForce RTX 4090 in terms of twice as much memory - it also has 1,792 more CUDA (shader) cores than the latter with 18,176. Compared to the old A6000, the difference is even more drastic - it only has 10,752 ALU/CUDA cores.
However, the new RTX 6000 is probably also more than twice as expensive as the RTX 4090 (around 1,950 Euros) - Nvidia hasn&t announced a price yet, but the old RTX A6000 currently costs from 5,690 Euros.
Nvidia RTX 6000
Nvidia doesn&t yet give any information about the clock rate, which is important for performance, but since a maximum power consumption of 300 watts is specified for the RTX 6000 and 450 watts for the RTX 4090, it can be assumed that the RTX 6000 is clocked significantly lower (its dual-slot DHE cooling system, which is adopted from the predecessor model, is also significantly smaller than the 3-slot cooler of the RTX 4090).
In addition to the CUDA cores, the RTX 6000, which is intended for workstations in the areas of 3D graphics, AI and video, has 142 third-generation RT (RayTracing) cores and 568 fourth-generation Tensor cores, also significantly more than the previous model with 84 RT and 336 Tensor cores. As with the predecessor, four DisplayPoerts 1.4 interfaces are on board.
The professional graphics card specialist PNY already has a page online about RTX 6000, but no information about price and release date yet.