Anthropic Ups Compute Deal with Google and Broadcom Amid Skyrocketing Demand

Why It Matters

The added compute capacity positions Anthropic to meet exploding AI demand, reinforcing its competitive edge against rivals like OpenAI and Microsoft while anchoring U.S. AI supply chains.

Key Takeaways

• Anthropic expands compute partnership with Google and Broadcom
• Adds 3.5 GW of AI compute capacity
• Leverages Google Cloud TPUs for Claude models
• $50 billion U.S. data‑center investment continues
• Run‑rate revenue jumps to $30 billion

Pulse Analysis

Anthropic’s latest compute deal underscores a broader industry shift toward massive, specialized hardware to sustain the rapid scaling of generative AI models. By securing additional Tensor Processing Units from Google Cloud and integrating Broadcom’s silicon, the company ensures that Claude can handle the exponential growth in query volume and model complexity. This strategic alignment mirrors moves by peers such as OpenAI, which also rely on custom chips and cloud partnerships to keep latency low and throughput high, a critical factor for enterprise adoption.

The agreement adds roughly 3.5 gigawatts of compute, a figure comparable to the power consumption of a small city, and will be deployed primarily in U.S. data centers. This expansion dovetails with Anthropic’s $50 billion pledge to bolster domestic AI infrastructure, a response to heightened geopolitical scrutiny and supply‑chain concerns. The timeline targets 2027, giving the firm a runway to refine Claude’s next‑generation models while capitalizing on Broadcom’s advanced networking and interconnect technologies, which promise lower latency and higher bandwidth for large‑scale training workloads.

Financially, the compute boost comes on the heels of a $30 billion Series G round that lifted Anthropic’s valuation to $380 billion and propelled its run‑rate revenue from $9 billion at the end of 2025 to $30 billion today. This growth reflects robust enterprise demand, with over 1,000 customers each spending more than $1 million annually. The expanded hardware capacity not only safeguards Anthropic’s market share but also positions it as a key player in the U.S. AI ecosystem, potentially influencing future policy decisions around AI safety and national security.