NVIDIA is currently cooking its next-gen AI GPUs in the oven right now, where we're hearing about the unannounced B200 AI GPU from... Dell.

NVIDIA has its beefed-up H200 AI GPU coming soon, while its next-gen B100 AI GPU should be revealed at GTC 2024 in less than two weeks time, but now it looks like we should expect a beefed-up B200 AI GPU in the future thanks to comments from Dell's Chief Operating Officer and Vice Chairman, Jeff Clarke.
Clarke confirmed that their engineering teams will be ready for a monster NVIDIA B200 AI GPU and that the next-gen AI GPU will use up to 1000W of power for a single GPU. As it stands, the Hopper H100 AI GPU will max out at around 700W of power for the SXM variant of the card; while we don't know power consumption numbers for B100 just yet, it should be north of 700W.
- Read more: NVIDIA's next-next-gen R100 GPU could be named after astronomer Vera Rubin
- Read more: NVIDIA's next-gen GB200 'Blackwell' GPU listed on its 2024 data center roadmap
- Read more: NVIDIA next-gen B100 AI GPU 'supply constrained, demand far exceeds supply'
- Read more: NVIDIA: next-gen B100 Blackwell GPU: over 4x as fast as H100 AI GPU
Dell COO and Vice Chairman Jeff Clarke said: "We're excited about what happens at the B100 and the B200, and we think that's where there's actually another opportunity to distinguish engineering confidence. Our characterization in the thermal side, you really don't need direct liquid cooling to get to the energy density of 1000 watts per GPU".
"That happens next year with the B200. The opportunity for us really to showcase our engineering and how fast we can move and the work that we've done as an industry leader to bring our expertise to make liquid cooling perform at scale".




