NVIDIA has developed a professional visualization and local AI solution called the RTX PRO 6000 Blackwell GPU, which is equivalent to the "Blackwell" Titan. The RTX PRO 6000 is based on the GB202 silicon and is the closest option to NVIDIA's prosumer-oriented Blackwell SKUs. It features 24,064 CUDA cores, just 512 cores short of the complete 24,576-core configuration of the GB202. This slight difference is likely due to yield defects, making it possibly the highest CUDA core count model based on GB202. The RTX PRO 6000 Blackwell also has 2,340 more cores than the top-end consumer GeForce RTX 5090, which has 21,760 CUDA cores. With a 600 W TGP, the RTX PRO 6000 Blackwell is expected to have lower clocks to accommodate its larger core count and massive memory configuration, as GDDR7 memory modules also consume power.
Featuring 96 GB of GDDR7 memory and ECC memory correction, the RTX PRO 6000 Blackwell is designed for professionals in 3D rendering, simulations, and local AI development. To cool this powerful configuration, NVIDIA has chosen a double-flow-through cooler similar to the one used on the RTX 5090, with an open-air design. The GPU operates on a PCIe 5.0 x16 interface with four DisplayPort 2.1 connectors and maintains a double-slot thickness. Pictures of the RTX PRO 6000 Blackwell have been obtained by VideoCardz, which can be seen below. The new cooler features darker color accents. More information about NVIDIA and the RTX PRO 6000 Blackwell GPU is expected to be revealed at the upcoming GTC 2025 event on March 17.
