Intel’s turnaround effort accelerated under new CEO Lip-Bu Tan, who returned to the company after resigning from its board and immediately faced tough internal questions about his departure and credibility. Political headwinds followed when Sen. Tom Cotton alleged extensive ties between Tan and Chinese tech firms, prompting a call from the president for Tan’s resignation. Days later, Tan met with the president and the two struck a rare agreement: Intel would transfer 9.9 percent of its stock to the federal government in exchange for funding, while Nvidia agreed to invest billions and partner on new chips that Intel will manufacture, with Nvidia committed to buying some output.
The stakes extend far beyond a corporate rescue. Leading edge chips used in defense systems and Artificial Intelligence are largely made by TSMC in Taiwan, with additional capacity in South Korea. The Pentagon relies on those offshore supplies, including for the F-35. Intel has not shipped true leading edge chips since 2017 but retains the fabs, tools, and know-how to attempt a return with its 18A and 14A nodes, with 14A not expected before 2027 at the earliest. Past missteps, such as declining early iPhone chip opportunities and failing to build a successful GPU while Nvidia surged, were compounded by CEO turnover and slipping performance. Pat Gelsinger secured a large CHIPS Act package and launched an ambitious fab buildout, but mounting costs and delays eroded investor confidence.
Tan is reversing course to stabilize finances and refocus execution. He has halted or canceled major factory projects, announced 15,000 layoffs to flatten 11 management layers, and is selling assets, including a majority stake in Altera and shares in Mobileye, with additional capital expected from strategic investors. Former CEO Craig Barrett and ex-director David Yoffie argue that deep-pocketed customers should inject new capital, a model Nvidia has now validated. Still, Bernstein’s Stacy Rasgon warns that everything hinges on Intel proving it can deliver 14A at scale, on spec, on time, and at acceptable cost. Intel’s own filings acknowledge that without at least one large 14A customer and key milestones, it may pause or discontinue that node and successors, a potential end to U.S. production of leading edge processors.
The rescue also tests industrial policy. Economists caution that government equity stakes often end poorly, raising questions about picking national champions. Yet national security experts counter that supporting Intel is about securing a domestic industrial base for strategically vital chips in an era of geopolitical competition. For Intel, the path forward blends fiscal discipline, hard engineering execution, and customer-backed financing. For the United States, the outcome could define supply chain resilience for Artificial Intelligence and defense for years to come.
