After a sharp drop in stock price and being considered for removal from the Dow Jones Industrial Average, Intel has suffered another setback. The company's latest 18A manufacturing process is reported to have failed Broadcom's test, making the prospect of turning its foundry business profitable even more remote.
According to the latest report from Reuters, sources revealed that Broadcom processed silicon wafers through Intel's most advanced 18A manufacturing process during testing. Broadcom retrieved the silicon wafers from Intel last month. After its engineers and executives studied the test results, they concluded that this manufacturing process is not yet suitable for large-scale mass production.
A Broadcom spokesperson said the company is "evaluating Intel Foundry's products and services and has not reached a final conclusion yet." According to people familiar with the matter, Broadcom's engineers still have doubts about the feasibility of this process, which may be related to the number of defects on the wafer or chip quality. And being able to improve the yield of wafers is crucial for achieving large-scale production. At present, it is impossible to determine the cooperation relationship between the two parties or whether Broadcom has decided to withdraw from a potential production deal.
Intel responded in a statement: "Intel's 18A process is already in use, running well and with good production volume. We still plan to start large-scale production next year. Although the industry is highly interested in the 18A process, for policy reasons, we will not comment on specific customer conversations."
Intel's transformation plan is fraught with difficulties
Since CEO Pat Gelsinger took the helm at Intel three years ago, the company has been presenting itself as a foundry. This is one of Pat Gelsinger's important strategies to turn the company around, seeking to regain a foothold in the high-end chip manufacturing field by manufacturing chips designed by other companies. However, in recent years, Intel has always lagged behind its competitors in the chip manufacturing field. The company's filing with the US Securities and Exchange Commission in April showed that its chip foundry business had an operating loss of $7 billion in 2023. However, Intel executives expect the foundry business to break even in 2027.
The test failure is another blow to Intel's highly anticipated foundry business, especially at a time when the company's overall business is facing challenges. Earlier in August, Intel reported second-quarter sales that fell 1% year-on-year to $12.8 billion, lower than analysts' expectations of $12.95 billion; adjusted earnings per share were $0.02, lower than analysts' expectations of $0.10. In addition, the company expects third-quarter sales to reach $12.5 billion to $13.5 billion, far below analysts' expectations of $14.38 billion; it expects an adjusted loss of 3 cents per share, while analysts expect earnings of 30 cents per share.
Intel also announced it will lay off more than 15,000 people and suspend dividends that have been in place since 1992. The company hopes to advance its operational model transformation and reverse the long-term loss situation of its chip manufacturing business. CEO Pat Gelsinger said the company plans to cut expenses by $10 billion in 2025, so "we must adjust the cost structure to conform to the new operational model."
In addition, it is worth mentioning that Lip-Bu Tan, a veteran in the semiconductor industry, stepped down from Intel's (INTC.US) board of directors at the end of August. The reason is that he has differences with the company's CEO Pat Gelsinger and other directors on issues such as Intel's redundant staff size, risk-averse culture, and lagging artificial intelligence strategy.
According to information, as part of the plan to restore Intel's status as a leading global chip manufacturer, Lip-Bu Tan joined Intel's board of directors two years ago. The company's board of directors expanded Lip-Bu Tan's responsibilities in October 2023 and authorized him to oversee the manufacturing business. Intel's board chairman Omar Ishrak once pointed out that Lip-Bu Tan is a highly respected global leader in the semiconductor industry. His expertise in software, semiconductors, and venture capital, deep industry relationships, and rich experience on public company boards will bring more valuable perspectives to Intel's board. However, this appointment ended abruptly in less than two years, showing the uncertainty of Intel's transformation plan.
Intel's investors have lost patience with the company's transformation plan, which involves heavy investment in new chips and production technologies while Intel is trying to do so amid shrinking revenues. More importantly, Intel's current products are not what those investing in artificial intelligence infrastructure want, which means Intel is losing market share.
Later this month, Pat Gelsinger and some other key executives will submit a plan to the company's board of directors to divest unnecessary businesses and cut capital expenditures. This plan will include ideas on how to achieve overall cost reduction by selling businesses including the programmable chip division Altera. In addition, according to people familiar with the matter, the company is discussing various possible scenarios, including splitting its product design and manufacturing businesses and which factory projects may be canceled.
