Intel shares fell 28.58% on Oct. 30 in U.S. trading, the company’s worst single-day drop since 1974, erasing more than $140 billion in market value overnight. In after-hours trading, the stock remained near $65.
The broader semiconductor sector also sold off. SanDisk dropped 13.47%, Qualcomm lost 11.12%, AMD and Arm each fell more than 10%, Broadcom, Taiwan Semiconductor Manufacturing Co. and Micron Technology were down about 5%, and Nvidia closed 4.22% lower. Losses stretched across chip design, memory and wafer manufacturing, but Intel posted the steepest decline.
The move came as AI spending continues to climb. Intel’s latest third-quarter results showed strong growth in its data center and AI business, but the company’s fourth-quarter guidance came in well below Wall Street expectations, prompting investors to reassess its growth outlook.
AI demand is rising, but Intel cannot ship enough product
Intel reported third-quarter revenue of $16.5 billion, up 21% from a year earlier. Revenue from its Data Center and AI business, or DCAI, reached $7.3 billion, up 78% year over year. Client Computing and Physical AI, or CCPG, generated $8.1 billion, down 5%. Total product revenue came to $15.4 billion, up 22%.
That mix shows how central the data center business has become to Intel’s growth story. On the earnings call, Chief Executive Officer Chen Liwu said that as AI applications shift from model training toward inference and agents, CPUs are taking on a larger role in task orchestration, multi-step reasoning and persistent execution.
Intel said global server CPU shipments are expected to grow by more than 30% in both 2026 and 2027. The company added that growth could be faster if supply conditions improve.
Demand, though, is outpacing what Intel can deliver. The company said on the earnings call that customer demand for data center products continues to exceed current supply capacity by a significant margin, mainly because of limits in wafer manufacturing capacity and shortages in advanced packaging materials.
Intel is increasing wafer output on its Intel 3 process, but capacity expansion still takes time. At the same time, high-end organic substrates used in advanced packaging are in short supply. Even when wafers are completed, processors cannot be delivered on time if packaging capacity lags behind.
The company said it is working to raise wafer output, improve yields and cycle times, and coordinate with suppliers to increase substrate supply. Management expects those supply constraints could last into 2027, leaving Intel unable to fully meet customer demand over the next several quarters.
That leaves Intel in a difficult position. AI-driven server demand is growing, but limited supply means part of that demand cannot be converted into revenue quickly. In a highly competitive server chip market, delivery capacity also affects Intel’s ability to win customers and defend market share.
Pressure increased further after an incident in mid-October at Intel’s wafer fab in Kiryat Gat, Israel, which halted production for about two weeks and disrupted server processor supply. Chief Financial Officer David Zinsner said the incident is expected to reduce fourth-quarter revenue by about $300 million and cut gross margin by about 200 basis points. Revenue in the first quarter of 2027 could face a hit of similar size. The plant is expected to gradually resume production by the end of October, but the impact may continue into the first half of 2027.
AI is pulling memory supply away from PCs
While Intel’s server business is constrained by production, its traditional PC segment is facing a different problem. Large AI data centers require large volumes of high-bandwidth memory, or HBM, and server DRAM. Because AI-related products carry higher margins, memory makers are shifting more capacity and resources toward the data center market, squeezing supply for traditional PCs.
Higher memory prices hit computer manufacturers first. Bill-of-materials costs rise, putting upward pressure on PC prices, while enterprise and consumer replacement budgets have not increased at the same pace.
Chen said on the earnings call that customers have already started adjusting PC purchasing plans based on memory pricing and supply conditions. As end-product prices rise, OEMs are accelerating inventory adjustments and reducing processor purchases.
Intel expects global PC shipments to decline by about 12% in 2026, with the drop in the second half of the year exceeding 20%. It expects another decline of about 10% in 2027. If memory supply and pricing remain under pressure, any recovery in the PC market could be delayed.
The effect is already visible in Intel’s numbers. In the third quarter, CCPG revenue was $8.1 billion, down 5% year over year and about 9% lower from the prior quarter. Intel expects revenue from that business to fall by more than 20% sequentially in the fourth quarter.
Zinsner said weak PC demand is only part of the issue. OEM customers are also actively cutting inventory to align with 2027 purchasing needs. Inventory adjustments are expected to remain a headwind from the fourth quarter through the first quarter of 2027.
Intel is now dealing with two pressures at once: strong AI server demand that it cannot fully serve because of supply limits, and a traditional PC market where customers are becoming more cautious as memory costs rise. Data center growth is strong, but the PC business remains a major source of revenue, and Intel cannot solve memory pricing and supply issues on its own.
18A is in production, but foundry still lacks a major outside customer
Another challenge sits in Intel’s manufacturing transition. The company has spent the past several years pushing its IDM 2.0 strategy in an effort to rebuild manufacturing competitiveness through advanced process technology and foundry services. Intel 18A is a key part of that plan.
At the start of 2026, Intel began mass production of products built on the 18A process, using RibbonFET gate-all-around transistors and PowerVia backside power delivery. Management said client CPU revenue tied to 18A has risen from less than 5% in the first quarter to nearly one-third, and related server products have already started shipping.
Still, internal production does not mean the external foundry business has broken through. For now, 18A is still mainly serving Intel’s own products, and the company has not announced any outside customer that has entered large-scale production on the node. Chen said Intel is working with potential customers on 18A-P, an enhanced version of 18A, and that customers could announce related developments over the next two quarters.
The next-generation Intel 14A process is also moving through customer engagement. In October this year, Intel released version 0.9 of the 14A process design kit, or PDK, and expects to launch version 1.0 in the first quarter of 2027. Large-scale production for customer products, however, is still some distance away.
The financial picture shows why investors remain cautious. Intel’s foundry business posted third-quarter revenue of $6.7 billion, up 58% year over year, but most of that came from internal transactions. Excluding those internal transactions, external foundry revenue was only $411 million, while operating loss for the period still reached $1.9 billion.
For Intel, the question of whether advanced process technology can reach production has a partial answer. Whether it can attract enough outside customers and turn the foundry business profitable remains unresolved.
Spending keeps rising while returns remain distant
Intel’s advanced process and foundry push has yet to generate enough return, but the company still needs to spend more. Fab expansion, equipment purchases and additional advanced packaging capacity all require continued investment.
Intel expects capital expenditures of about $20 billion for full-year 2026, with a sharp increase again in 2027. Chen said on the earnings call that the company is accelerating equipment purchases, and formal purchase orders sent to suppliers rose nearly 40% from the previous quarter.
Near-term revenue and profit growth are still not keeping pace with that spending. Intel forecast fourth-quarter revenue of $15 billion to $16 billion, below Wall Street’s prior expectation of about $17 billion. It expects non-GAAP gross margin of 38%, down 5 percentage points from the third quarter, and non-GAAP earnings per share of $0.25, below analysts’ expectation of $0.43.
Cash flow is under pressure as well. Intel expects full-year 2026 free cash flow to remain negative. As of the end of the third quarter, the company held about $21.84 billion in cash and cash equivalents and had about $46.32 billion in long-term debt. Intel has previously raised funds through equity financing, but future capital spending will still require sustained financial support.
Investors had spent the past year rebuilding confidence in Intel’s manufacturing transition. As capital spending keeps climbing, attention is shifting to a harder question: when those investments will turn into revenue, profit and cash flow. AI has opened a new growth path for Intel, but the company still needs time to benefit from it in full, and market patience is being tested.
