Given its $6,800 price tag, you probably won't see Nvidia RTX 6000 Ada on any lists of the best graphics cards for gaming. But no one can deny that the RTX 6000 Ada is a powerhouse for professional users and content creators. The RTX 6000 Ada Generation, as it's officially known, features Nvidia's flagship AD102 silicon, which the chipmaker also uses in the mainstream GeForce RTX 4090. For the RTX 6000 Ada, Nvidia enabled 142 SMs out of the possible 144, so the graphics card is just two SM shy of using the full AD102 die. The RTX 6000 Ada arrives with 18,176 CUDA cores, 568 Tensor cores for AI workloads, and 142 RT cores for ray tracing tasks. With the potential to boost up to 2,505 MHz, the graphics card flaunts an impressive FP32 performance that tops out at 91.06 TFLOPs. There's also 48GB of 20 Gbps GDDR6 memory onboard with a 384-bit memory interface to land the graphics card at a memory bandwidth of up to 960 GBps. Note that the TGP (Total Graphics Power) on the 6000 Ada is capped at 300W, however, compared to 450W for the RTX 4090 Founders Edition.
The previous RTX A6000 delivered 38.71 TFLOPs FP32 performance and 768 GBps of memory bandwidth. Therefore, the RTX 6000 Ada offers 135% higher FP32 performance and 25% more memory bandwidth than the RTX A6000. The GeForce RTX 4090, with 82.58 TFLOPs FP32 performance and 1,008 GBps memory bandwidth, is probably the closest rival to the RTX 6000 Ada.