Intel resumes stalled US chip plant expansions amid AI demand surge

Progress on stalled US chip plants

Intel has resumed work on two major US semiconductor manufacturing sites, signaling a shift in its expansion strategy as demand for AI and data center processors grows. The company is installing equipment at Fab 62, its new chip plant near Phoenix, Arizona, with operations expected to begin in early 2028. Fab 62 is adjacent to Fab 52, which is already operational on Intel’s Ocotillo campus.

In Ohio, where progress had stalled, Intel has installed the first major roof truss at its delayed chipmaking site. The Ohio facility, once envisioned as the world’s largest chipmaking hub, is now slated to open between 2030 and 2031, pushed back from earlier timelines.

Demand drives spending and delays

Intel’s manufacturing chief, Naga Chandrasekaran, and CEO Lip-Bu Tan have linked the resumption of these projects to surging demand for the company’s data center processors, which power AI services. Orders for these chips are reportedly outpacing Intel’s production capacity.

The company has raised its spending plans to over $20 billion for 2026 and expects even higher expenditures in 2027. Additionally, Intel is investing $5.7 billion to expand production of its Xeon chips in Ireland.

Despite these investments, Intel has yet to announce major customers for its chipmaking-for-others business. Reports from May suggested Apple was in early talks to use Intel’s factories, but no formal agreements have been disclosed. Intel has also joined Elon Musk’s Terafab project as a manufacturing partner.

Government ties and competitive pressures

Intel’s expansion efforts coincide with its financial strategies, including a proposed $15 billion share sale in August. The US government holds a stake in the company, reflecting broader national interests in semiconductor independence.

Intel’s spending remains below that of competitors like TSMC, which is expected to invest around $60 billion this year. The delays in Ohio and the cautious approach to Fab 62 underscore the challenges of balancing capacity expansion with demand certainty.

What this means

Lazyfounder analysis — our interpretation, not reported fact.

Intel’s renewed progress on Fab 62 and Ohio signals a bet on long-term demand for AI and data center chips, but the delays and staggered timelines reveal a cautious, demand-driven approach. For founders and operators in the semiconductor and AI industries, this highlights three key dynamics:

1. Capacity as a Competitive Lever: Intel’s ability to scale production—even with delays—could reshape supply chains for AI infrastructure. Competitors like TSMC and Samsung are already outspending Intel, but the Ohio site’s eventual capacity (if realized) could become a critical asset for US-based manufacturing. Operators in AI hardware or cloud services may need to factor Intel’s expanded capacity into their sourcing strategies by 2030.

2. Government Stakes Raise Stakes for Execution: With the US government as a stakeholder, Intel’s projects are no longer just commercial ventures but also geopolitical tools. Founders in adjacent sectors (e.g., AI infrastructure, defense tech) should monitor how this influences Intel’s priorities, including potential subsidies, export controls, or partnerships tied to national security.

3. Customer Commitments Remain a Weak Spot: Intel’s chipmaking-for-others business lacks marquee customers, unlike TSMC’s roster (which includes Apple). The reported Apple talks suggest Intel is trying to close this gap, but until it secures high-volume clients, its foundry ambitions remain uncertain. For startup founders building AI hardware, Intel’s progress may offer an alternative to TSMC—if the price, capacity, and reliability align. For now, the lack of confirmed deals signals that Intel has ground to make up.

Related on Lazyfounder

• More technology news