By Max A. Cherni, Jeffrey Dastin and Stephen Nelis
The new CEO of San Francisco (Reuters) -intel (Intc) is exploring a major change in contracting business to win big customers, two people familiar with the question said in a potentially expensive shift from their predecessor plans.
If implemented, the new strategy for what Intel calls its “foundry” business will not lead to the fact that it no longer offers on the market a certain technology for the production of chips, which the company has long developed, to external customers, people said.
Ever since he took over the company’s rudder in March, CEO Lip-Bu Tan has moved quickly to reduce costs and find a new path to revive a sick American chipmaker. By June, he began to express that the production process, which previous CEO, Pat Gel -Gelsinger, was betting strongly, known as 18A, was losing his appeal to new customers, the sources said, who spoke on condition of anonymity.
In order to put aside external sales of 18A and its 18A-P version, production processes that cost Intel billions of dollars to develop, the company will have to take a write-off, said one of the people familiar with the question. Reuters -related industry analysts said such a fee could be a loss of hundreds of millions, if not billions, dollars.
Intel declined to comment on such “hypothetical scenarios or market speculation”. It states that the leading client for 18A has long been Intel himself and aims to launch the production of his Panther Lake chips in 2025, which he calls the most modern processors ever designed and manufactured in the United States.
The persuasion of external customers to use Intel factories remains key to its future. As its 18A production process has been confronted with delays, TSMC’s N2 technology is on its way to manufacture.
TAN’s preliminary response to this challenge: Focus more resources on 14A, the next -generation chip process, in which Intel expects to have advantages over TSMC of Taiwan, the two sources said. This move is part of a play for big customers such as Apple and NVIDIA, who are currently paying TSMC to produce their chips.
File Photo: Intel Executive Director, LIP-BU TAN, speaks in California
Tang has instructed the company to strengthen the discussion options with the Intel board when it meets this month, including whether to stop marketing 18A by new customers, one of the two sources said. The Council may not decide on 18a until the subsequent autumn meeting in the light of the complexity of matter and the huge money, the person said.
Intel declined to comment on what he calls hearing. She said: “Lip-buu and the executive team are committed to strengthening our roadmap, building confidence with our customers and improving their financial situation for the future. We have identified clear areas of focus and will take the action necessary to turn the business.”
Last year was the first unprofitable year of Intel in 1986. It reported a net loss due to the $ 18.8 billion company for 2024.
Intel CEO reports show the huge risks – and costs – considered to move the historic American chipmaker back to a solid basis. Like Gelsinger, Tan inherited a company that has lost its production advantage and lags behind the decisive technological waves of the last two decades: mobile calculations and artificial intelligence.
The company focuses on large-volume production later for 18A with its internal chips, which are expected to arrive before the external customer orders. Meanwhile, delivering 14A on time to win large contracts is by no means certain and Intel can choose to stick to its existing plans for 18A, the sources said.
Intel adjusts 14A to the key needs of customers to make it successful, the company said.
Amazon and Microsoft of 18A
Tan’s review of focusing customers on 14a includes the part of an Intel or foundry chips contract, which makes chips for external customers.
Regardless of the board’s decision, Intel will make chips through 18A in cases where his plans are already on the move, people familiar with the question said. This includes the use of 18A for Intel’s internal chips that it has already designed for this production process, people say.
Intel will also produce a relatively small volume of chips, which it guarantees for Amazon.com and Microsoft in 18A, with deadlines that make it unrealistic to wait for the development of 14A.
Amazon and Microsoft did not immediately comment on the question. Intel said he would fulfill his clients’ commitments.
Tan’s overall strategy for Intel remains nascent. So far, he has updated his leadership team, attracting new engineering talents and working to shrink what he considers to be an inflated and slow -moving average management.
A detachment from the sale of 18A to the foundry customers will be one of its largest moves so far.
The 18A production process involves a new method of delivering energy to chips and a new type of transistor. Together, these improvements were intended to allow Intel to match or exceed the capabilities of TSMC, Intel leaders said earlier.
According to some industrial analysts, however, the 18A process is approximately equivalent to the so -called TSMC N3 production technology, which was produced in large volume production at the end of 2022.
If Intel follows the TAN management, the company will focus its foundry employees, designer partners and new clients of 14a, where he hopes for a bigger chance to compete against TSMC.
Tan is based on extensive contacts and relationships with customers built over decades in the chip industry to reach its view of 18A, the two sources said.
(Report from Jeffrey Dastin, Max A. Cherni and Stephen Nelis in San Francisco; Edit by Kenneth Li and Matthew Lewis)
