Intel (INTC.US) is under mounting pressure to finally get its struggling foundry business off the ground, following a record single-day stock plunge driven by a deteriorating PC market outlook.
The worsening PC market has sent Intel shares tumbling by the largest single-day margin since the 1970s. For Intel, the foundry business has been slow to gain traction. Over the past five years, the company has poured a significant portion of its $108 billion in capital expenditures into an attempt to challenge TSMC (TSM.US), yet it has still not secured major customers for its most advanced chips and remains primarily engaged in manufacturing chips for its own products.
Wall Street is losing patience. After the U.S. government acquired a 10% stake in Intel in August 2025, Nvidia (NVDA.US) followed with a $5 billion investment, sending Intel's stock soaring to six times its original value within 10 months and peaking in June. Intel had finally found its place in the AI boom, supplying central processing units for agentic AI, while optimism about its chip fabrication vision was also heating up. But last Thursday, Intel's situation took a sharp turn for the worse. The company said quarterly revenue would fall $1 billion to $2 billion short of analyst expectations, as the PC market continues to be battered by surging costs for memory and other components. The PC business still accounts for roughly half of Intel's revenue, and consumers are increasingly reluctant to upgrade their devices. On Friday, Intel shares plummeted 29%, the worst single-day performance since the 1970s; according to FactSet data, the stock fell 45% for the month of October, the largest monthly decline on record. In Monday trading, shares dropped another 8.8%.
Intel's investments in its future should be starting to pay off. CEO Lip-Bu Tan said the global server processor market could expand by as much as fivefold over the next five years, reaching approximately $200 billion. In the third quarter, Intel's data center business grew 78% year-over-year to $7.3 billion. However, Intel said it cannot produce enough chips to meet demand, citing industry-wide supply constraints in wafers, substrates, and memory, many of which are also hurting PC sales.
Five Years of Investment Still Without Major Customers, 18A Capacity Forced to Shift to Older Nodes
This has forced Tan to make difficult choices. To address the capacity shortage, Intel said it is shifting some production at its Oregon facility from advanced 18A chip wafers to the older Intel 3 node for its data center products. Intel is counting on 18A to attract external foundry customers, so the resource shift could complicate that effort. Harvard Business School professor David Yoffie, who served on Intel's board for nearly three decades, said the latest move is hardly welcoming to potential foundry customers. "They worry that if Intel doesn't actually have excess capacity beyond its own needs, then they're not a very credible supplier," Yoffie said. If a large-scale foundry customer is not found within a year, "the pressure on Intel's economics will become more apparent."
An Intel spokesperson said in a statement: "Customer confidence in Intel Foundry is built on execution. We are delivering on our technology roadmap, demonstrating strong performance on 18A and progress on 14A. We continue to see strong customer engagement and are focused on earning business through consistent execution." Intel is best known for its industry-leading central processing units (CPUs). Nvidia and AMD (AMD.US) graphics processing units (GPUs) dominated the early AI boom because of their excellence in handling large workloads for model training and chatbot inference; Intel's CPUs are essential for running the virtual machines used by agentic applications like Meta (META.US) Muse. Mizuho Securities analyst Jordan Klein said in an interview that as one of the few companies globally capable of manufacturing and packaging advanced chips for its own products, Intel can "control its own destiny" better than most chipmakers. "If you don't have enough capacity, you're leaving revenue and money on the table," Klein said.
Former CEO Pat Gelsinger, who was ousted at the end of 2024, aggressively pursued foundry investments during his four-year tenure to catch up after delays in 7-nanometer and 10-nanometer technologies. These advanced process nodes are designed to pack more transistors onto each chip. Tan has taken a more cautious approach, limiting spending until customer demand is locked in. He said last Thursday that "several customers" would be announced over the next two quarters. Intel's own 18A chips are in high-volume production in Arizona, and a second facility is under construction. The first structural truss has just been installed at the new Ohio facility, but the project is nearly five years behind schedule.
Both Competitor and Foundry: Customer Concerns and Spinoff Debate
Yoffie said another obstacle to Intel attracting new customers is that chipmakers like Broadcom (AVGO.US) and Qualcomm (QCOM.US) may be reluctant to manufacture chips at a company that is simultaneously a competitor on the product side. "The right solution is for Intel to spin off its foundry business," Yoffie said, adding that this is "not only the best option, but perhaps the only option." Klein said AMD may be the only major chipmaker not interested, since the two companies compete directly in the CPU market. He called it a "competitive, confidentiality, and privacy issue." AMD is set to report third-quarter earnings after the bell on Tuesday. Yoffie said the United States needs Intel to succeed. "Ultimately, TSMC can build at most 20% to 25% of its global capacity in the U.S., and that will still take years," he said. "If the U.S. doesn't have its own foundry, once any geopolitical issue arises or a supply chain problem like during COVID occurs, that would be bad for the country and bad for the U.S. semiconductor industry."
Comments