Broadcom jumps as long-term Google TPU and Anthropic compute deals extend to 2031
Broadcom shares are rising after the company disclosed long-term AI infrastructure agreements tied to Google’s next-generation TPUs and expanded support for Anthropic’s TPU-based compute buildout. The deals extend supply visibility into 2031 and include roughly 3.5 gigawatts of TPU compute capacity for Anthropic starting in 2027.
1. What’s driving AVGO higher today
Broadcom (AVGO) is moving higher as investors react to newly disclosed, multi-year AI infrastructure agreements that deepen its role in Google’s TPU roadmap and expand its collaboration with Anthropic. The disclosures point to extended duration and clearer volume visibility for custom AI accelerators and related data-center components, which traders are treating as a bullish signal for Broadcom’s AI-led growth trajectory. (finance.yahoo.com)
2. The key details investors are focusing on
The agreements include a long-term commitment with Google to develop and supply future generations of custom Tensor Processing Units, plus a supply assurance arrangement for networking and other components used in Google’s next-generation AI racks, extending through up to 2031. Separately, the expanded Anthropic collaboration calls for Anthropic to access about 3.5 gigawatts of next-generation TPU-based AI compute capacity through Broadcom beginning in 2027—an unusually concrete data point for future demand. (finance.yahoo.com)
3. Why the market cares right now
Broadcom is a critical supplier to the “picks-and-shovels” layer of AI—custom silicon, high-speed connectivity, and rack-level components—and multi-year agreements can reduce uncertainty around the pacing of AI infrastructure spend. With the TPU ecosystem scaling, the market is pricing in stronger, longer-duration demand for Broadcom’s custom AI silicon and adjacent platform content, helping explain the outsized upside move versus the broader tape. (tomshardware.com)
4. What to watch next
Investors will look for follow-through in order visibility, customer concentration disclosures, and any incremental commentary on AI revenue trajectory and profitability as custom silicon ramps. The next potential catalysts are management updates on AI backlog/commitments and any additional multi-year capacity or supply agreements with hyperscalers and large AI model developers that mirror the Google/Anthropic structure. (finance.yahoo.com)