Back to list
Meta and Broadcom Extend Strategic AI Chip Partnership Through 2029 as Hock Tan Exits Board
Industry NewsMetaBroadcomArtificial Intelligence

Meta and Broadcom Extend Strategic AI Chip Partnership Through 2029 as Hock Tan Exits Board

Meta has officially extended its collaborative agreement with Broadcom for the development of AI chips, securing a partnership that will now run through 2029. This extension underscores the ongoing technical synergy between the social media giant and the semiconductor leader. Alongside this strategic renewal, Meta disclosed in a recent filing that Broadcom CEO Hock Tan will not be standing for reelection to Meta's board of directors. Tan, who joined the board in 2024, informed the company of his decision last week. This leadership shift occurs even as the two companies deepen their long-term commercial ties in the competitive artificial intelligence hardware sector.

Tech in Asia

Key Takeaways

  • Partnership Extension: Meta and Broadcom have formally extended their AI chip deal, ensuring collaboration until 2029.
  • Board Resignation: Broadcom CEO Hock Tan has announced he will not seek reelection to Meta’s board of directors.
  • Short Tenure: Tan’s departure from the board comes after joining the leadership team in 2024.
  • Strategic Continuity: Despite the board change, the long-term technical agreement between the two firms remains intact and expanded.

In-Depth Analysis

Extension of the Meta-Broadcom AI Chip Deal

Meta has solidified its long-term hardware strategy by extending its existing AI chip agreement with Broadcom. According to recent disclosures, the partnership is now set to continue through 2029. This extension is a significant indicator of Meta's reliance on Broadcom's expertise to facilitate its custom silicon ambitions. As Meta continues to scale its infrastructure to support increasingly complex artificial intelligence models, the stability provided by a multi-year deal with a primary semiconductor partner is critical for their hardware roadmap.

Leadership Changes: Hock Tan’s Departure from Meta’s Board

While the commercial relationship between the two companies is strengthening, the governance structure is seeing a notable shift. Meta revealed in a filing that Broadcom CEO Hock Tan informed the company last week of his intention not to stand for reelection to the board. Tan’s tenure on the Meta board was relatively brief, having only joined in 2024. The filing did not provide specific reasons for his decision to step down, but it highlights a transition in the personal leadership overlap between the two tech giants even as their corporate interests remain closely aligned through 2029.

Industry Impact

The extension of the deal between Meta and Broadcom through 2029 signals a period of sustained investment in custom AI hardware. For the broader industry, this move highlights the importance of securing long-term supply chains and design partnerships in the face of global chip demand. The departure of Hock Tan from the board, while a significant change in corporate governance, appears to be decoupled from the technical and commercial success of the partnership, suggesting that the operational collaboration is robust enough to transcend individual board seats.

Frequently Asked Questions

Question: How long is the new AI chip deal between Meta and Broadcom?

The agreement has been extended to run through the year 2029.

Question: Why is Hock Tan leaving the Meta board of directors?

According to Meta's filing, Hock Tan informed the company he will not stand for reelection. The specific reasons for this decision were not detailed in the original report.

Question: When did Hock Tan originally join Meta's board?

Hock Tan joined the Meta board of directors in 2024.

Related News

Meta Removes Commercial Terms Language Permitting AI Content Scraping Across Publisher and News Websites
Industry News

Meta Removes Commercial Terms Language Permitting AI Content Scraping Across Publisher and News Websites

Following widespread public and industry backlash, Meta has updated its commercial terms to remove language that indicated the company could scrape third-party web content, including material from news organizations, to train its artificial intelligence models. The revision specifically strikes provisions from its commercial terms documentation that previously asserted Meta could utilize content across associated domains, web pages, and URLs to develop, improve, research, train, and maintain its products and services. The policy change comes amid mounting scrutiny over how artificial intelligence developers acquire training data from online publishers. This update reflects the sensitive balance tech companies must navigate between gathering web-scale data and addressing copyright, consent, and commercial boundaries asserted by content creators and digital publishers.

Websites Are Closing Their Doors to AI Agents as Autonomous Web Scraping Sparks Friction Across the Internet
Industry News

Websites Are Closing Their Doors to AI Agents as Autonomous Web Scraping Sparks Friction Across the Internet

Autonomous AI agents have become the central focus of major technology developers, with OpenAI, Meta, Google, Instinct, and other companies pledging that their systems will soon manage everyday digital tasks autonomously. These promised capabilities include scouring the web on behalf of users, completing online shoe purchases, booking dining reservations, and organizing cluttered email inboxes. However, this emerging paradigm faces severe friction as digital platforms increasingly resist automated interaction. As websites begin closing their doors rather than cooperating with autonomous programs, a fundamental question emerges regarding what happens when the broader internet refuses to cooperate. This standoff highlights a growing divide between AI developers seeking unhindered web access and web platforms acting to protect their digital boundaries, user environments, and proprietary online interfaces.

US AI Coding Startup Replit Ends Work With Anthropic as Enterprise Revenue Expands to 15 Percent
Industry News

US AI Coding Startup Replit Ends Work With Anthropic as Enterprise Revenue Expands to 15 Percent

United States-based artificial intelligence coding platform Replit has officially ended its working relationship with frontier AI model developer Anthropic, according to a report published by Tech in Asia. The decision marks a noteworthy operational development for the coding startup as it navigates an increasingly competitive landscape for generative software engineering tools. Concurrently, the company has achieved substantial growth in its commercial client base, with enterprise contracts rising to account for approximately 15% of Replit's total revenue this year, compared to just 2% recorded in the prior year. While specific operational reasons and transition details behind the termination were not elaborated upon in the initial reporting, the shift reflects broader structural adjustments as developer platforms scale institutional sales and manage external model dependencies.