At this year’s Siggraph event, NVIDIA’s Jensen Huang spoke to Wired for an hour about NVIDIA, RTX, and AI. Among other things, he said that artificial intelligence systems will require three times as many computers in the future.
The NVIDIA CEO noted that the world of artificial intelligence is now moving away from its initial phase and towards the next one, which he called the «corporate wave». After that comes the «physical wave». He clarified that three computers will be needed: one to create AI, another to model and improve AI, and the third — to run the AI itself.
«It’s a three-computer problem. You know, the three-body problem, and it’s so incredibly complex, and we built three computers to do it».
Huang was referring to a number of NVIDIA hardware and software packages, ranging from DGX H100 servers for creating AI, Jetsen embedded computers for simulating AI, and workstations and servers that use Omniverse and RTX GPUs to run AI.
Do you really need three computers for AI? Even companies that really want to integrate AI into their core business can’t help but think about the potential cost and complexity of using and paying for three tiers of NVIDIA products. The hardware manufacturer is now gets the most out of artificial intelligence technologies and wants to do so in the future, so such statements should be treated critically.
Source: PC Gamer