There’s also software improvements to consider, there’s a lot of room for efficiency improvements.
Why does that make a difference? Compute for AI is build on the progress for compute first for GPU then for data center. They are similar in nature.
fidodo@lemmy.world 4 months ago
ryannathans@aussie.zone 4 months ago
Building an ASIC for purpose built computation is significantly faster than generic gpu compute cores. Like when ASICs were built for bitcoin mining/sha256 and a little 5 watt usb device could outperform the best GPUs
frezik@midwest.social 4 months ago
It may be even more specialized than that. It might be a return to analog computers.
Which isn’t going to work for Nvidia’s traditional products, either.
Buffalox@lemmy.world 4 months ago
The H200 is evolved from Nvidia GPU designs, and will be by far the most powerful AI component in existence when it arrives later this year, AI is now so complex, that it doesn’t really make sense to call it an ASIC, and the cost is $40,000.- so no not small 5 watt units.