Anthropic expands Google + Broadcom partnership for multiple gigawatts of next-generation compute — Claude's training capacity now structurally above one-time procurement
Anthropic announced expansion of its compute partnership with Google and Broadcom for multiple gigawatts of next-generation accelerator capacity. The deal locks in training and inference capacity across multiple years and includes custom-silicon allocation — Anthropic's first material commitment to TPU/Broadcom-designed accelerators outside the AWS Trainium relationship. The financial scale is consistent with the $965B Anthropic valuation backdrop.
The substantive piece is the supply-side commitment that backs the demand-side scaling. Anthropic at $965B valuation needs the compute capacity for the Claude Code revenue trajectory the market is pricing in. Multi-gigawatt commitments — historically the territory of hyperscaler capex announcements — are now a frontier-lab procurement pattern as Anthropic, OpenAI, and Google all secure 3-5 year supply visibility.
The strategic frame is that Anthropic now has explicit compute-supplier diversification: AWS Trainium (long-running), Google TPU (this announcement), and Broadcom custom silicon (this announcement). That diversification is the structural protection against any single supplier's capacity constraints or pricing shifts. For OpenAI — increasingly Azure-bound for the same buyer-base — the Anthropic supplier-diversification posture is a competitive advantage on supply resilience.
Anthropic — Anthropic expands partnership with Google and Broadcom for multiple gigawatts of next-generation compute → · CNBC — Microsoft and Google take on Anthropic and OpenAI in AI coding models → · AI Business — OpenAI vs. Anthropic vs. Google: But the Model Isn't the Point →