Lam Research Sees Sold-Out Chip Demand, Targets Capacity and AI Service Gains

Lam Research NASDAQ: LRCX said it is prioritizing customer execution, capacity readiness and product-roadmap delivery as semiconductor equipment demand remains strong and customers seek tools sooner than the company can provide them. During a Goldman Sachs discussion with Senior Equity Analyst Jim Schneider, an unnamed Lam executive said the company is focused on meeting customer commitments, improving on-time delivery, managing lead times and quality, and expanding hiring to support new fabs in multiple regions. The executive added that Lam is working to ensure near-term operational demands do not distract from its longer-term research-and-development and product-delivery plans.

The executive said the semiconductor industry is “fundamentally sold out” and constrained by available clean-room capacity. Lam sees between eight and 10 new clean rooms at its tier-one customers coming online between now and the end of next year. Customer discussions have intensified around immediate tool needs, requirements over the coming year and expected demand beyond that period, according to the executive. Lam is using those forecasts to plan its manufacturing footprint, hiring, installation teams and supplier requirements.

While acknowledging that the semiconductor industry has historically been cyclical, the executive said Lam’s more immediate concern is missing upside demand rather than preparing for a downturn. The company said it has developed an operating model over four decades to manage potential downturns, but that is not its primary focus at present.

On the large investments being made for artificial intelligence infrastructure, the executive said end customers see an opportunity to build barriers to entry through compute infrastructure. “If you’re not doing it, you’re going to be at a fundamental competitive disadvantage to someone that is,” the executive said of AI adoption.

Lam is applying AI internally to help field engineers install, troubleshoot and service its equipment. The company is training AI models on more than 30 years of documented troubleshooting history, aiming to shorten the time required to resolve equipment issues and improve engineer productivity. The executive said roughly two-thirds of Lam’s field engineering community is using the capability. Beyond productivity gains, Lam sees an opportunity to use such tools to support advanced service offerings. Lam also said it is expanding resources near customers as new U.S. fabs are developed. The company recently announced a lar...