Company city, United States
San Francisco Bay Area
Where the AI labs, the chip start-ups and many neoclouds are run from, and home to the El Capitan and Perlmutter supercomputers.
The Bay Area is the main customer and company base of AI compute rather than its largest site, since power and land are expensive. OpenAI, Anthropic and most frontier labs are here, and so are the chip companies selling their own clouds, Cerebras, Groq and SambaNova. Neoclouds such as Lambda, Together AI and Crusoe are run from the area, and two national lab supercomputers sit at its edge.
Chips sold as clouds
Cerebras in Sunnyvale, Groq in Mountain View and SambaNova in Palo Alto all began by selling AI chips and systems. Each now sells access to its own hardware as a cloud service, mostly for inference, the running of trained models, by the token. Cerebras listed on the Nasdaq in May 2026. In December 2025 Nvidia licensed Groq's technology and hired its founder, while GroqCloud kept running.
Lab machines
Lawrence Livermore National Laboratory, east of the bay, runs El Capitan, an AMD-based exascale machine that became the fastest on the TOP500 list in November 2024. Lawrence Berkeley National Laboratory runs Perlmutter at its NERSC centre, an Nvidia A100 system that serves thousands of researchers. These are used for science and weapons work but have been opened to AI research too.
Providers of San Francisco Bay Area
Still to be written: Google Cloud Platform.
Also tied to San Francisco Bay Area
Photo credits on this page
The small pictures above come from Wikimedia Commons under free licences. Some are cropped to fit.
- Cerebras: Coolcaesar, CC BY 4.0, source
- Cloudflare: HaeB, CC BY-SA 4.0, source
- Groq: Avivweinstein, CC BY-SA 4.0, source