Broadcom is widening its AI push with two new partnerships that tie the company more tightly to the next phase of chip design and manufacturing. The company, together with Meta, UCLA and other partners, is launching a US$125 million AI semiconductor research and talent hub over five years, while also working with Applied Materials on next generation AI chip packaging technologies.
Broadcom shares were trading at about US$414.57, after a year in which the stock returned 81.3%. The five-year gain was close to 9x, and the shares were up 19.3% year to date, even after falling 5.7% over the past week and rising 3.1% over the past month. For investors watching avgo stock, the new agreements show a company trying to extend its advantage beyond demand for AI hardware and into the research base and manufacturing know-how that support it.
The UCLA hub puts Broadcom directly into early stage AI semiconductor research and talent development. It also gives the company another route into the academic side of the field at a time when the competition for engineers and specialized expertise has become part of the AI race itself. The Applied Materials partnership sits closer to manufacturing, with access to Applied’s global EPIC centers potentially helping Broadcom design advanced packaging techniques for future AI chips.
That matters because the push into energy efficient, high performance AI computing depends on more than faster logic alone. Packaging has become one of the most important choke points in the industry, and advanced packaging research can demand heavy capital and long development cycles before it turns into commercial products. Broadcom is not just buying time in that race; it is building relationships at both ends of it, from university labs to manufacturing facilities.
The partnerships also reinforce Broadcom’s role in custom accelerators and Ethernet switching for hyperscalers such as Alphabet, Nvidia and AMD. Those businesses sit at the center of the AI buildout, where customers want more compute, lower power use and systems that can move data quickly enough to keep massive clusters running efficiently. The new hub and the Applied Materials work suggest Broadcom is trying to make itself harder to displace in that chain.
The near-term question is whether these deals translate into products and talent that deepen Broadcom’s lead, or whether they remain strategic investments that take years to pay off. For now, the answer looks clear: Broadcom is betting that the companies that control the best AI infrastructure, and the know-how behind it, will be the ones that keep the edge.
