The Information’s Amir Efrati and Kalley Huang say they have learned that Meta Platforms (META) is putting the final touches on a 100,000 GPU cluster, which will be a little bigger than 100,000 H100s, located somewhere in the U.S. The company will use the new supercomputer cluster to train the next version of its Llama model – Llama 4, according to a person familiar with the matter. The cost of the chips alone could be more than $2B, the authors note. While the report does not say who is providing the Ethernet switches, analysts at Evercore ISI think “there is a high probability” that it is Arista Networks (ANET).
Elevate Your Investing Strategy:
- Take advantage of TipRanks Premium at 50% off! Unlock powerful investing tools, advanced data, and expert analyst insights to help you invest with confidence.
Published first on TheFly – the ultimate source for real-time, market-moving breaking financial news. Try Now>>
Read More on META:
- Evercore sees ‘high probability’ Meta 100K+ cluster networked by Arista
- Moderna, Dow slide after offering sales outlooks: Morning Buzz
- Amazon and Meta: Top Analyst Chooses the Best Internet Stocks to Buy
- Facebook blocked emergency warnings during wildfires, Washington Post reports
- Meta admits to scraping Australian users’ public data to train AI, The ABC says