Intel’s crews installed the first major structural truss at the company’s chip fabrication plant in New Albany, Ohio. That’s no small thing. The milestone starts the next phase of construction on a project running at least five years behind schedule. The delay isn’t a rounding error; it’s the kind of gap that makes investors and customers ask hard questions.
Intel Foundry head Naga Chandrasekaran cast the moment in familiar corporate terms. “This represents another significant step toward expanded U.S. leading-edge semiconductor capacity,” Chandrasekaran wrote on the company’s website Wednesday. The sentiment is fine. The timeline is the problem.
The Timeline Slips
Intel announced the Ohio factory in 2022, four years ago as of today. The company originally expected to begin delivering chips in 2025, roughly a year from now in the calendar sense, but that date is gone. Intel now aims to finish construction in 2030. The company plans a second Ohio fab for completion in 2031 at the site, just northeast of Columbus.
That’s a long wait for a plant that was supposed to be online by now. Taxpayers, shareholders and the communities that bet on the project have all been asked to wait, and patience has a cost.
The slowdown reflects a broader recalibration under CEO Lip-Bu Tan, who took the helm in 2025. The Ohio factory forms part of Intel’s effort to make chips for outside customers, a build-out begun under former CEO Pat Gelsinger. Tan’s turnaround strategy took a more cautious approach: wait for customer demand before moving ahead with ambitious fab plans. That’s a sober choice. Building capacity nobody has ordered is how companies go broke.
Demand Is the Real Test
Intel still needs customers to make the investment pay off. In September, reports said SK Hynix, the world’s leading maker of high-bandwidth memory, could partner with Intel in Ohio to lease foundry space for memory-chip production at the new site. Neither company confirmed plans. Until someone signs, the factory is a bet, not a business.
The Arizona operation offers a more concrete picture. By late 2025, Intel had started high-volume production at its Arizona chip fab, making its most advanced 18A chips. Chandrasekaran said Wednesday that operations at Intel’s next Arizona fab are expected to begin by the end of 2028. Yet Intel still hasn’t secured a major external customer for the Arizona-made 18A chips, and the company uses them for its own processors. That’s an internal supply chain, not a market verdict.
Intel’s history explains the caution. Once a dominant chip manufacturer, the company suffered major delays with its 7- and 10-nanometer nodes, contributing to the stock’s worst year ever in 2024, two years ago. Customers have been reluctant to commit again since. Trust, once broken in manufacturing, takes years to rebuild.
A Competitive Opening
Intel may now have an opportunity amid record AI demand because Taiwan Semiconductor Manufacturing Co. can’t make chips quickly enough. Competition for production space on TSMC’s factory lines is “extremely fierce,” even for Nvidia, Advanced Micro Devices, Apple and Broadcom, according to Mizuho Securities analyst Jordan Klein.
That’s where Intel hopes to compete. TechInsights analyst Dan Hutcheson sees the logic plainly. “People want a second source to be able to twist TSMC’s arm and keep them honest on pricing,” Hutcheson said. “All the big ones have been testing there.”
Competition keeps prices honest, and a second source is exactly what a market needs. Whether Intel can deliver on that promise, though, depends on execution and customers, not on the trusses going up in Ohio.
Why This Matters:
The Ohio fab tests whether private capital and corporate discipline can rebuild American chip manufacturing without open-ended reliance on public money or political promises. Intel’s delay, now stretching at least five years past its original target, shows how quickly capital-intensive projects can drift when demand is uncertain. The company has pulled back to wait for customers, a market-driven choice that avoids stranding capacity but leaves Intel dependent on outside buyers it has yet to sign. Any signed commitment from a firm such as SK Hynix would validate the investment, while continued silence would raise hard questions about the return on tens of billions of dollars in construction. For policymakers and shareholders alike, the lesson is that a building going up isn’t the same thing as a business taking shape.
- intel
- ohio chip fab
- new albany
- semiconductor manufacturing
- chip factory construction
- chip demand
- intel foundry
- sk hynix