The new CEO of exclusive enterprise explores major transformation in chip manufacturing business
By Max A. Cherney, Jeffrey Dastin and Stephen Nellis
Two people familiar with the matter told Reuters that the new CEO of San Francisco (Reuters) is exploring major changes in his contract manufacturing business to potentially be expensive compared to his predecessor’s plans.
If implemented, people say, what Intel calls the “casting” business will no longer market some of the chip manufacturing technologies the company has long developed for external customers.
Since seizing the company’s helm in March, CEO Lip-Bu Tan has quickly cut costs and found new ways to recover sick U.S. chipmakers. Sources said that by June, he began expressing the manufacturing process, which Pat Gelsinger bets a large number of bets (18A) are losing appeal to new customers.
One familiar with the matter said that in order to put aside the external sales of 18A and its variants of 18A-P, the manufacturing process cost billions of dollars in development, the company must promote. Industry analysts contacted by Reuters said the allegations could cost hundreds of millions of dollars (or even billions of dollars) of losses.
Intel declined to comment on such “supposed scenarios or market speculation.” It says the main customer of 18A has long been Intel itself, and it aims to gradually produce its “Panther Lake” laptop chip later in 2025, which it calls it the most advanced processor ever made in the United States.
Convincing external customers to use Intel’s factory remains the key to its future. As its 18A manufacturing process faces delays, competitor TSMC’s N2 technology is already on track.
Tan’s preliminary answer to this challenge: Focusing more resources on the 14A is Intel’s next-generation chip manufacturing process, in which Intel expects an advantage over Taiwan’s TSMC. The move is part of major customers such as Apple and Nvidia, who are currently paying TSMC to make chips.
Tan’s mission is an option the company will discuss with Intel’s board of directors at its earliest meeting this month, including whether to stop marketing 18A to new customers, one of the two sources said. The person said the board may not make a decision on the 18A until the subsequent fall meeting, given the complexity of the matter and the huge amount of money.
Intel declined to comment on the alleged rumors. “Lip-Bu and the executive team are committed to strengthening our roadmap, building trust with customers, and improving future financials. We have identified clear areas of focus and will take the actions needed to change the business,” it said in a statement.
Last year was Intel’s first unprofitable year since 1986. Its net loss was attributed to $18.8 billion in 2024.
Intel CEO’s review shows the huge risks and costs of considering returning Story Us Chipmaker to a solid foothold. Like Gelsinger, Tan inherits a company that has lost its manufacturing advantage and lags behind the critical technology wave of the past two decades: mobile computing and artificial intelligence.
The company’s internal chips will target 18A later this year, and its internal chips will be widely expected to arrive before external customer orders. Meanwhile, there is no doubt that 14A will be delivered in time to win a major contract, and Intel can choose to stick to its existing 18A plan, a source said.
The company said Intel is tailoring the 14A for the successful quantification of key customers.
Amazon and Microsoft 18A
Tan’s comment on whether to focus customers on 14A involves Intel or foundry contracts that make chips for external customers.
No matter what the board’s decision is, Intel will make chips through 18A if the plan is already in the plan, people familiar with the matter said. This includes using 18A internal chip for Intel, which has been designed for the manufacturing process, people say.
Intel will also produce relatively few chips and guarantee that Amazon.com and Microsoft will pass 18A chips, with a deadline that makes it unrealistic to wait for 14A development.
Amazon and Microsoft did not immediately comment on the matter. Intel said it will fulfill its customer commitments.
Tan’s overall strategy for Intel is still very fresh. So far, he has updated his leadership team, introduced new engineering talents, and worked to narrow down what he believes is swollen and slow intermediate management.
The transfer from sales of 18A to casting clients will represent one of his biggest moves to date.
The 18A manufacturing process includes a novel method that can deliver energy to a chip and a novel transistor. Intel executives have said before that these enhancements are designed to match or surpass TSMC’s capabilities.
However, according to some industry analysts, the 18A process roughly corresponds to what TSMC calls N3 manufacturing technology, which is engaged in mass production in late 2022.
If Intel follows Tan’s lead, the company will focus on foundry employees, design partners and new customers on 14A, hoping for a better opportunity to compete with TSMC.
Tan draws on the extensive connections and customer relationships that have been established in the chip industry over decades to reach the 18A view, two sources said.
(Reported by Jeffrey Dastin, Max A. Cherney and Stephen Nellis in San Francisco; Editors of Kenneth Li and Matthew Lewis)
