Lex Fridman Podcast
#459 – DeepSeek, China, OpenAI, NVIDIA, xAI, TSMC, Stargate, and AI Megaclusters
Nathan Lambert
And then likewise, if you fill out that entire 1.8 gigawatts with the next two generations of NVIDIA's chips, GB200, GB300, VR200, and you fill it out completely, that ends up being roughly $50 billion server cost. Plus there's data center costs, plus maintenance costs, plus operation costs, plus all these things. And that's where OpenAI gets to their $100 billion announcement that they had.
0
💬
0
Comments
Log in to comment.
There are no comments yet.