Talk about scale—xAI's infrastructure setup is getting wild. A $400 million Tesla Megapack energy storage deployment handles the power demands, while GPU inventory hits over $30 billion once fully operational. The third facility they acquired will push total training compute to nearly 2GW capacity, establishing what's shaping up to be the world's most powerful training infrastructure. The engineering ambition here is hard to overstate.
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
5 Likes
Reward
5
5
Repost
Share
Comment
0/400
MeltdownSurvivalist
· 4h ago
Wow, 2GW of computing power? How much electricity would that cost? Haha
View OriginalReply0
SchrodingerAirdrop
· 4h ago
My goodness, this scale is really outrageous—$40 billion worth of GPU inventory? Just pouring money into building computing power directly.
View OriginalReply0
ConfusedWhale
· 4h ago
Wow, this infrastructure scale is really incredible, 2GW of training compute power. Elon Musk is really throwing money this time.
View OriginalReply0
SudoRm-RfWallet/
· 4h ago
Wow, 2GW of training capacity? This is going to the sky!
View OriginalReply0
ETH_Maxi_Taxi
· 4h ago
Wow, this infrastructure scale is really outrageous. Elon is starting to burn money again to pile up hardware...
Talk about scale—xAI's infrastructure setup is getting wild. A $400 million Tesla Megapack energy storage deployment handles the power demands, while GPU inventory hits over $30 billion once fully operational. The third facility they acquired will push total training compute to nearly 2GW capacity, establishing what's shaping up to be the world's most powerful training infrastructure. The engineering ambition here is hard to overstate.