Skip to content

broadcom google

Here’s how Google and Anthropic are turbocharging Broadcom

Broadcom shares are rising after the company struck a deal with Google to supply AI chips and expand its collaboration with Anthropic, the developer of Claude. The American microchip manufacturer is thus attempting to challenge Nvidia's dominance, the global leader in AI semiconductors. All the details.

Broadcom shares jump forward thanks to long-term contracts for AI chips, rivals to Nvidia’s dominant GPUs, the largest semiconductor group in the world.

Yesterday the American microchip manufacturer signed a long-term agreement with Google for the production of future versions of the tech giant’s artificial intelligence chips. The agreement indeed provides that Broadcom will develop and supply custom tensor processing units for Google’s next-generation TPUs.

At the same time, Broadcom announced an expansion of its collaboration with Anthropic, developer of the AI model Claude. The chip company will provide the AI startup led by Dario Amodei access to about 3.5 gigawatts of AI computing capacity, leveraging Google’s AI processors, starting from 2027.

These initiatives strengthen the strategic collaboration among the three companies. Although the companies have not disclosed the financial terms of the agreements, Broadcom shares rose about 3% in after-hours trading.

As Reuters recalls, demand for custom chips like Google’s tensor processing units (TPUs), used for AI workloads, has skyrocketed in recent years as companies seek alternatives to Nvidia’s expensive graphics processors. Meanwhile, Anthropic’s popularity has surged this year, with its Claude app becoming the most downloaded free app in the United States on Apple’s App Store in February, after a controversy between the company and the Pentagon became public, notes Cnbc.

Therefore, summarizes the Financial Times, Anthropic will spend hundreds of billions of dollars on Google chips and cloud services, with part of the capacity provided through the partnership with Broadcom.

All the details.

WHAT THE AGREEMENT WITH GOOGLE PROVIDES FOR…

Thus Broadcom will develop chips that use Google’s tensor processing units (TPUs), offering an alternative to Nvidia’s technology. As Bloomberg, explains, the Alphabet-controlled tech company designed the TPUs to speed up its ubiquitous search engine, but they have proven useful for creating and running artificial intelligence software. Broadcom takes Google’s specifications and creates complete designs that can then be sent to production.

Additionally, a supply assurance agreement will guarantee that the chip company led by Hock Tan will provide network components and other parts for Google’s future AI racks through 2031.

AND THE EXTENSION OF THE PARTNERSHIP WITH ANTHROPIC

In this context, Anthropic, Broadcom, and Google are expanding a strategic collaboration that will allow Anthropic to access about 3.5 gigawatts of computing power starting in 2027.

During a conference call on last month’s financial results, Broadcom CEO Hock Tan stated that “for Anthropic, we had a great start in 2026” by providing 1 gigawatt of computing power from Google’s proprietary tensor processing units (TPUs). “For 2027, this demand is expected to increase to exceed 3 gigawatts of computing power,” he added.

Hock Tan had also said that Broadcom expects sales of its AI chips to exceed $100 billion next year, thus becoming a more significant competitor to Nvidia.

HOW MUCH THE DEAL WITH AMODEI’S AI COMPANY WEIGHS FOR THE CHIPMAKER

In a note following the earnings call, Mizuho analysts, led by Vijay Rakesh, estimated that Broadcom would generate $21 billion in revenue from Anthropic’s AI in 2026 and $42 billion in 2027, CNBC notes again.

SUPPORTING THE CLAUDE BOOM

Meanwhile, yesterday Anthropic stated that the new agreement with Broadcom builds on the company’s commitment to invest $50 billion to strengthen U.S. computing infrastructure.

“This innovative partnership with Google and Broadcom represents the continuation of our rigorous approach to infrastructure scalability: we are building the capacity needed to handle the exponential growth we have seen in our customer base, while enabling Claude to define the frontier of AI development,” said Krishna Rao, Anthropic’s CFO, in a blog post. Most of the new infrastructure will be located in the United States, Anthropic said.

After all, demand for its AI model Claude increased in 2026, with over 1,000 enterprise customers spending more than $1 million annually. This figure has more than doubled since February. As demonstrated by Anthropic’s annual revenue increase, which surpassed $30 billion, compared to about $9 billion at the end of last year.

 

 

 

Back To Top