Vibeleaderboard
← All Intel
Intel / video

Jensen Huang – Will Nvidia’s moat persist?

Source
youtube.com
Author
Dwarkesh Patel
Date

Context

Dwarkesh Patel asked NVIDIA chief executive Jensen Huang, in an interview, whether AI could commoditize NVIDIA the way many expect it to commoditize software, since NVIDIA designs chips that TSMC and others manufacture. Huang answered that the company's job is turning electrons into (the units of text a model reads and writes), that doing this well is hard to copy, and that NVIDIA does as much as necessary and as little as possible, partnering for the rest.

Huang also argues software will gain from AI: more agents mean more instances of tools such as chip-design software, held back today because agents are not yet good at using them. On supply, Patel noted NVIDIA's purchase commitments, which Patel put near $100 billion in filings. Huang says suppliers invest because NVIDIA's downstream demand is large enough to absorb their output, and that bottlenecks such as advanced packaging get cleared once demand is clear. These are the claims of an executive whose company sells the hardware.

Terms in this piece · Glossary
  • token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
  • inference — Running a trained model to get answers — the phase where AI is actually used, as opposed to trained.
Read the source www.youtube.com
More from Dwarkesh Patel
Recommended reads
Comments

Checking sign-in…

Loading comments…