On October 29 local time, U.S. chip giant Intel announced its third-quarter 2026 financial results. Driven by strong growth in its Data Center and AI (DCAI) business, the company’s quarterly revenue and adjusted earnings per share both exceeded prior guidance, with data center revenue increasing 78% year-over-year—the strongest growth in 15 years.
However, Intel's guidance for the fourth quarter fell below market expectations. The company also stated that total capital expenditures for 2026 are expected to remain at $20 billion, with a significant increase planned for 2027 to expand manufacturing capacity and meet demand for AI server CPUs.
Intel CFO David Zinsner said 2027 will be an investment year for the company, during which positive free cash flow may not be achieved. Following weaker guidance, an incident at its Israel facility, and weak demand in the PC market, Intel’s stock dropped 23% in after-hours trading after the earnings report.
PC market drags on short-term performance
The financial report showed that Intel's revenue for the third quarter reached $16.513 billion, a 21% year-over-year increase; net income attributable to Intel was $3.3 billion, a 18% decline year-over-year. On a non-GAAP basis, adjusted net income was approximately $2.8 billion, up 169% year-over-year; adjusted earnings per share were $0.52, compared to $0.23 in the same period last year. Intel stated that this marks the eighth consecutive quarter in which revenue, gross margin, and earnings per share exceeded prior guidance.
Looking at individual businesses, data center and AI operations emerged as the most significant growth drivers this quarter. DCAI revenue reached $7.318 billion, representing a 78% year-over-year increase and a 17% sequential increase; operating profit amounted to approximately $3 billion, accounting for 42% of DCAI revenue. Chief Financial Officer Zinsner stated that the third-quarter results further underscore the critical role of server CPUs in AI infrastructure development.
In comparison, client computing and physical AI business (CCPG) revenue was $8.097 billion, a 5% year-over-year decline. AI PC revenue increased by 35% year-over-year, but this was insufficient to offset the impact of overall weakness in the PC market. Intel’s foundry business generated $6.697 billion in revenue, a 58% year-over-year increase, but the segment still recorded an operating loss of $1.853 billion for the quarter, indicating it remains far from sustained profitability.
In terms of profitability, Intel's non-GAAP gross margin for the third quarter was 44.1%, an increase of 4.1 percentage points year-over-year. Zinsner attributed the improvement to lower costs from the Intel 3 and Intel 18A processes, as well as product mix optimization. The company generated $5.1 billion in operating cash flow for the quarter.
However, what truly disappointed the market was the guidance for next quarter's performance.
Intel expects fourth-quarter revenue of $15 billion to $16 billion, non-GAAP earnings per share of $0.25, and a non-GAAP gross margin of 38%. In comparison, LSEG’s previous market consensus estimated revenue of approximately $17.07 billion and adjusted earnings per share of $0.43. Both revenue and profit guidance from the company are significantly below market expectations.
Before the earnings report, Intel's stock closed at $92.07 during regular trading hours on October 29, down 3.62%. After the earnings release, the stock plunged approximately 23% in after-hours trading.
In addition, Intel faces financial pressure due to production constraints, a weakening PC market, and ongoing investments in its manufacturing operations. Specifically, although the company has seen strong demand growth for AI server CPUs, it remains unable to fully meet customer needs in the short term, while rising memory prices continue to weigh on its client PC business.
Capacity and memory shortages form a dual constraint.
During the subsequent earnings call, Intel CEO Lip-Bu Tan linked the growth of its data center business to demand for AI infrastructure. He noted that market demand in the third quarter was more balanced compared to previous quarters, with strong momentum in data center and AI operations driving results above expectations. In particular, as AI agent applications evolve, CPUs are gaining new sources of market demand.
“Even with conservative estimates of the total number of digital agents and the number of agents served by a single CPU, it is easy to foresee a massive potential market size for CPUs in terms of sales and shipments over the next five years,” said Chen Liwu.
Zinsner further provided industry insights during the conference call, stating that the supply shortage in the server CPU market, driven by AI demand, could persist through 2027 and extend into 2028.
Intel is aiming to meet this demand with its next-generation products. Chen Liwu stated that Granite Rapids, based on the Intel 3 process, has become one of the company’s fastest-scaling server products, and the first server product based on the Intel 18A process, Clearwater Forest, has begun shipping and recognized revenue. The company expects strong demand for these products in 2027, particularly for high-density CPUs and agent-based rack-scale infrastructure.
Purpose-built silicon is also a key focus area for Intel. Chen Liwu stated that Intel continues to see strong momentum in purpose-built silicon, with revenue increasing by more than 50% quarter-over-quarter and more than doubling year-over-year. This indicates Intel’s intent to further expand its business in purpose-built silicon, advanced packaging, and external foundry services beyond its traditional general-purpose CPUs.
However, increased server CPU demand does not mean Intel can immediately convert market demand into revenue. Chen Liwu acknowledged that the company is currently facing manufacturing supply constraints. In addition to limited capacity in Intel 3 process technology, a shortage of high-end organic substrates required for advanced packaging is also hindering Intel’s ability to scale up server CPU production.
Another source of pressure comes from the PC market.
Chen Liwu stated that in recent weeks, customers have begun signaling a slowdown in the PC market based on memory prices and supply conditions. Zinsner expects global PC shipments to decline by approximately 12% in 2026, with a drop of more than 20% in the second half of the year; PC shipments may also fall by around 10% in 2027. He noted that the PC market will struggle to achieve sustained recovery until memory supply improves and prices return to normal. Inventory adjustments are also intensifying short-term pressures, with Intel expecting its client business revenue to decline by more than 20% quarter-over-quarter in the fourth quarter.
Additionally, an incident at Intel’s facility in Israel in October caused a production halt of approximately two weeks. Zinsner stated that preliminary estimates indicate the event will reduce fourth-quarter revenue by approximately $300 million and depress gross margin by about 2 percentage points, with a similarly sized impact potentially extending into the first quarter. The incident also affected the shipment of Intel’s 10-nanometer server products in the fourth quarter. The company is currently evaluating whether the wafers on the production line can still be used.
In terms of manufacturing processes, Intel has released some positive signals. Chen Liwu stated that the yield rates for both the Intel 3 and Intel 18A processes continued to improve in the third quarter, with Intel 18A achieving double-digit sequential growth in yield. Wafer output for both processes exceeded expectations. Intel 18A-P remains on track for use in internal products by year-end, and Intel 14A is progressing as planned; a version 0.9 process design kit (PDK) was released to select customers in October, with the company noting that this version met industry-standard yield and performance metrics at launch.
Zinsner stated that 2027 will be Intel’s year of investment. As Intel 18A and Intel 3 processes ramp up, and preparations begin for the production launch of Intel 14A in 2028, the company will need to make multiple investments in the same year, leading to an expectation that positive free cash flow may not be achieved in 2027. Intel aims to convert the demand generated by AI infrastructure into long-term revenue and profits through sustained investment in expanding capacity, improving product yields, and enhancing manufacturing efficiency.