Intel’s New CEO Plots Course Correction with Focus on AI and Manufacturing
SAN FRANCISCO, March 17 – Intel’s incoming CEO, Lip-Bu Tan, is poised to implement major shifts in the company’s strategy, including a renewed emphasis on artificial intelligence and a revamped approach to chip manufacturing, according to sources familiar with his thinking. This comprehensive plan aims to revitalize the tech giant after a period of challenges.
Tan, who is returning to the CEO role, is expected to address what he sees as inefficiencies, including a slow-moving and bloated middle management structure. Revamping Intel’s manufacturing arm, known as Intel Foundry, is also a core priority, the sources said. The foundry, once solely dedicated to Intel’s chip production, has expanded to manufacturing semiconductors for external clients.
At a town hall meeting following his appointment as CEO, Tan indicated the need for “tough decisions,” according to two people briefed on the meeting.
Revitalizing Intel Foundry
In the short term, Tan intends to enhance the performance of Intel Foundry, which produces chips for companies like Microsoft and Amazon. This will involve aggressively pursuing new customers. Plans include restarting efforts to produce chips for AI servers and focusing on areas beyond servers, such as software, robotics, and AI foundation models.
“Lip-Bu will be spending a lot of time listening to customers, partners, and employees as he comes on board and works closely with our leadership team to position the business for future success,” an Intel spokesman said in a prepared statement.
Strategic Shifts
Tan’s strategy appears to be, at least initially, a refinement of the approach taken by outgoing CEO Pat Gelsinger. Gelsinger’s plan focused on transforming Intel into a contract chip manufacturer to compete with Taiwan Semiconductor Manufacturing Co. (TSMC). Gelsinger committed billions of dollars to build factories in the U.S. and Europe.
Tan has been a vocal internal critic of Gelsinger’s initiatives.
Intel reported an annual loss of $19 billion in 2024, its first since 1986.
Overcoming Past Weaknesses
For most of its history, Intel exclusively manufactured chips for itself. However, under Gelsinger, the company’s manufacturing efforts for external clients fell short of the service and technical support provided by TSMC, leading to delays and failed tests.
Months of reviewing Intel’s manufacturing process, following his appointment to a special oversight role in late 2023, have informed Tan’s views. In his assessment, he expressed frustration with the company’s culture, believing that it had lost the “only the paranoid survive” ethos of former CEO Andy Grove. Decision-making was also seen as hampered by an excessive workforce, according to reports.
Tan presented some of his ideas to Intel’s board last year, but they were not implemented. He resigned abruptly in August over differences with the board. Now, returning as CEO, he will re-evaluate Intel’s workforce, which was reduced by roughly 15,000 employees to almost 109,000 at the end of last year. Tan has little choice but to make Intel’s existing manufacturing operations successful in the short term.
Future of AI Chips
Intel’s next generation advanced chips, incorporating AI features and known as Panther Lake, will rely on the company’s in-house factories using a new set of techniques and technologies it calls “18A.” Intel’s financial success this year depends on strong sales of this forthcoming chip.
Tan has indicated plans to retain control over the factories which remain financially and operationally separate from the design business and restore Intel’s position as a “world-class foundry.”
Industry analysts and Intel executives say that Intel’s contract manufacturing operation can succeed if Tan secures at least two major customers for high-volume chip production. Part of this effort will include improvements in its chip manufacturing process to make it easier for potential clients such as Nvidia and Google to use.
Intel has shown improvements in its manufacturing processes in recent weeks, which have sparked early test runs from Nvidia and Broadcom. Advanced Micro Devices is also evaluating Intel’s process.
Tan is expected to focus on enhancing output, or “yield”, with the goal of producing higher numbers of chips per silicon wafer. The aim is to move to an annual release schedule for AI chips, similar to Nvidia, but this will take years. Industry analysts and Intel sources believe it will be at least 2027 before Intel can develop a compelling new architecture for their first AI chip.
