Hock Tan built Broadcom by buying things. So it carried some weight when the chief executive, who assembled one of the industry’s largest chip companies through a long run of acquisitions, said dealmaking has slipped down his list of priorities.
Speaking at the Bloomberg Tech conference in San Francisco on Thursday, Tan said Broadcom is less focused on acquisitions now because AI offers stronger growth than an outside transaction is likely to.
The logic is arithmetic. With AI revenue surging, Tan questioned what acquisition could plausibly match the growth rate the company is already getting organically.
For a serial acquirer, the calculation has inverted: where buying scale once accelerated growth, the AI business is now growing fast enough on its own that a deal would more likely dilute the trajectory than lift it.
The 💜 of EU tech
The latest rumblings from the EU tech scene, a story from our wise ol’ founder Boris, and some questionable AI art. It’s free, every week, in your inbox. Sign up now!
That is a notable shift in posture for Broadcom specifically. The company’s identity has been bound up with its acquisitions, a strategy that turned a mid-tier chipmaker into a sprawling semiconductor and software conglomerate. Tan stepping back from that playbook, even rhetorically, signals how completely the AI cycle has rearranged the incentives for the companies supplying it.
Broadcom’s AI position is not abstract. The company has become a central player in custom AI silicon, acting as the design-and-supply layer between hyperscalers’ in-house chip ambitions and the workloads that run on them, including a long-term role in the Anthropic-Google compute arrangement built around custom TPUs and a seat in Google’s four-partner chip supply chain assembled to challenge Nvidia in inference. That business is the organic growth Tan is choosing over the acquisitions that defined his earlier strategy.
There is a practical dimension beneath the strategy too. Large acquisitions take years to clear regulators and integrate, and in a market moving as fast as AI silicon, that lag is itself a cost; capital tied up in a slow-closing deal is capital not spent on the custom-chip programmes that are compounding now.
Choosing organic growth is partly a bet that speed matters more than scale-by-purchase at this point in the cycle.
The comment also reads as a marker of where the cycle sits. When the dominant suppliers stop hunting for deals because their existing business is compounding faster than anything they could buy, it says something about how much value the AI build-out is concentrating in the incumbents.
Tan is not warning of a slowdown; he is describing growth so strong it makes the usual growth tactic redundant, the same surge of capital that has Google committing up to $40bn to Anthropic rather than competing with it head-on.
Whether the discipline holds is its own question. Acquirers tend to return to the habit when organic growth cools, and Tan’s framing is explicitly conditional on AI revenue continuing to surge.
For now, the most acquisitive figure in chips is sitting on his hands by choice, and saying the reason is that the thing he would buy growth with is already growing faster than the things he could buy.
