Broadcom’s CEO emphasizes the company’s focus on engineering excellence and strategic partnerships, such as with Google and Anthropic, to lead in the rapidly evolving AI chip market while encouraging internal adoption of AI tools to boost productivity. He highlights Broadcom’s collaborative yet competitive approach to custom AI silicon development, confidence in its networking leadership, and a preference for organic growth over large acquisitions amid the generative AI boom.
In the interview, Broadcom’s CEO discusses the current AI chip market and investor expectations, emphasizing the importance of focusing on fundamentals rather than stock price fluctuations amid the hype cycle. He acknowledges the rapidly evolving landscape, with frequent shifts in leadership and technology, and highlights Broadcom’s strategy of continuous investment and engineering excellence to stay ahead, even as customers like Google explore in-house chip design through “customer owned tooling.” Despite this, Broadcom competes by delivering differentiated products that outperform alternatives, particularly against dominant GPU technologies from Nvidia.
The CEO elaborates on Broadcom’s partnership with Google, especially regarding Google’s TPU chips, and how Google is working to democratize TPU access primarily through software and cloud services. Broadcom supports this effort but recognizes that software development remains largely under Google’s control. He also touches on Broadcom’s collaboration with AI foundation model builders like Anthropic, explaining that their investment was a calculated leap of faith based on the potential of generative AI, particularly in enterprise applications such as coding assistance. This partnership, alongside Google’s, has proven successful so far, with Anthropic showing strong growth and preparing for an IPO.
Regarding the use of AI tools internally, Broadcom encourages engineers to adopt AI-assisted design and coding tools without imposing strict usage limits. The CEO notes that productivity gains from these tools are significant, enabling senior engineers to accomplish in a week what previously took multiple engineers months to complete. He stresses the importance of learning to use these tools effectively to maximize their benefits and sees AI as a transformative productivity enhancer rather than a cost to be throttled excessively.
On the topic of Broadcom’s AI chip development with OpenAI, the CEO confirms that the partnership is progressing well, with production expected to begin later in the year. He dismisses rumors of complications or dependencies on Microsoft agreements, describing the collaboration as part of a broader journey where multiple customers are at different stages of developing custom silicon to optimize AI workloads. This reflects a competitive yet cooperative environment where Broadcom supports various AI leaders in their chip innovation efforts.
Finally, the CEO addresses Broadcom’s broader business strategy, including mergers and acquisitions, noting that the explosive growth driven by generative AI reduces the immediate need for large-scale acquisitions. He remains open to opportunities but finds it challenging to identify targets that match Broadcom’s rapid organic growth. He also comments on competition in networking, an area critical to AI infrastructure, expressing confidence in Broadcom’s market leadership and strategic customer focus. While competitors like Cisco are active, Broadcom’s extensive portfolio and sustained investment position it strongly to meet the massive demand for AI-related networking solutions.