Lam Research Lifts 2026 WFE Outlook as AI Demand Strains Fab Capacity

Lam Research raised its 2026 wafer-fabrication equipment spending outlook to the low-$150 billion range from $135 billion-$140 billion, driven by sustained AI demand and expectations for eight to 10 new leading-edge fabs through 2027. AI hardware is increasing equipment intensity, while advanced chip designs and packaging are expanding Lam’s addressable market. The company cited growing opportunities in gate-all-around transistors, backside power delivery, advanced packaging and dry-resist technology. Lam is accelerating capacity expansion, including a second Malaysia facility, as clean-room availability and supply-chain capacity constrain the industry. The company also expects NAND upgrades to accelerate, while targeting mid-50% gross margins and reporting record customer-support revenue. Lam Research NASDAQ: LRCX executives said the semiconductor equipment industry remains constrained by available clean-room capacity despite a higher outlook for wafer-fabrication equipment spending, citing sustained artificial intelligence demand and a growing pipeline of new fabs. Speaking at the Citi Global TMT Conference, Executive Vice President and Chief Financial Officer Doug Bettinger said Lam had raised its view for calendar 2026 wafer-fabrication equipment, or WFE, spending to the low-$150 billion range from an earlier estimate of $135 billion to $140 billion. “The industry is fundamentally under-supplying demand, and clean room is a constraint,” Bettinger said. He said customers have been able to increase output by using available fab space more efficiently, pulling projects forward and bringing existing clean rooms into production sooner than anticipated. Bettinger said Lam sees eight to 10 new tier-one fabs coming online through the end of 2027, supporting its expectation that WFE spending will grow next year. He added that the company’s top-down and bottom-up assessments point to continued AI demand as computing shifts from training workloads toward inference and eventually “Physical AI.” AI hardware raises equipment intensity Vice President of Investor Relations Ram Ganesh said Lam’s estimate that each $100 billion of AI data-center capital spending could generate roughly $8 billion of WFE demand has increased by $1 billion to $2 billion. He attributed the revision to a combination of higher hardware content, greater capital intensity and increased memory requirements. Ganesh said Lam has raised its NAND serviceable available market per wafer estimate to 2 times from 1.8 times presented at its analyst day. He also said device complexity has increased relative to the assumptions incorporated in the company’s original model. Bettinger said Lam is having unusually intensive planning discussions with customers and suppliers to ensure capacity is available. Lam is expanding its manufacturing capabilities, including accelerating work on a second facility in Malaysia and increasing output from existing facilities through denser use of production space. While supply-chain constraints can emerge because Lam’s tools contain thousands of parts, Bettinger characterized those issues as manageable. He said the company is working to ensure it is not a production bottleneck as customer demand rises. Foundry, logic and packaging opportunities Bettinger said Lam’s serviceable available market has expanded more quickly than expected. The company previously projected its addressable market would ...