What did Intel report for the third quarter of 2026?
Intel reported third-quarter revenue of $16.5 billion on 29 October 2026, up 21% from $13.7 billion a year earlier. Adjusted earnings per share came in at $0.52, against the $0.40 analysts polled by LSEG had expected, according to CNBC.
The headline numbers beat forecasts. Non-GAAP net income was $2.8 billion, up 169%, and non-GAAP gross margin reached 44.1%, up 4.1 percentage points. GAAP net income fell 18% to $3.3 billion, and GAAP earnings per share dropped 33% to $0.60.
The split inside the company matters more than the total. Data Center and AI revenue rose 78% to $7.3 billion. Chief financial officer David Zinsner called it Intel's strongest data center revenue growth in 15 years.
Client Computing and Physical AI Group, the unit that sells PC processors, generated $8.1 billion, down 5%. Intel Foundry revenue rose 58% to $6.7 billion. The smaller All Other category fell 30% to $0.7 billion, and intersegment eliminations removed $6.3 billion.
Intel reports across four lines. Intel Products combines the client and data center businesses and reported $15.4 billion, up 22%. Foundry builds wafers for Intel's own product groups and for outside customers. All Other covers smaller operations that sit outside those two core units.
The quarter showed how much operating leverage Intel has recovered. GAAP operating margin was 15.3%, up from 5.0% a year earlier. Research, development, marketing and general costs rose 4% to $4.5 billion, well below the 21% revenue increase.
Intel generated $5.1 billion of cash from operations in the quarter. It also completed a $23 billion public equity offering, announced plans to acquire quantum software company Classiq, and set out a multiyear Oracle Cloud Infrastructure collaboration that includes its next-generation Xeon processor, code-named Diamond Rapids.
Product announcements crowded the release. Intel said the first Googlebook devices on its Core Ultra Series 3 chips will come from Acer, ASUS and Lenovo. It also described a collaboration with SAP on what the company calls physical AI, linking business systems to equipment on factory floors.
Why did Intel's shares fall after the results?
Guidance, not the quarter, drove the selloff. Intel forecast fourth-quarter revenue of $15.0 billion to $16.0 billion. Analysts polled by LSEG expected $17.1 billion, CNBC reported.
The profit outlook was weaker still. Intel guided non-GAAP earnings to $0.25 a share and non-GAAP gross margin to 38.0%, down from $0.52 and 44.1% in the third quarter. GAAP earnings were guided to $0.10 a share.
At the midpoint, the revenue forecast is about 6% below the third quarter, and the gross margin guide sits 6.1 points lower than the figure Intel just reported. Both moves point in the same direction: less revenue at a lower profit rate.
Shares fell 16% in extended trading, according to CNBC. The stock had risen about 150% this year before the report and had already dropped roughly 23% in October on concerns about the scale of AI spending. Expectations had run well ahead of the numbers.
Zinsner gave two reasons for the cautious quarter. He said the company sees "the PC slowdown continuing at least through the first half of next year". Intel also lost output after an explosion at a fab in Israel killed two workers and seriously injured a third earlier in October. Zinsner told Reuters the incident would cost about $600 million in revenue over two quarters, and said production would ramp again by the end of the week.
Supply is the other constraint. Chief executive Lip-Bu Tan said his "immediate priority is to unlock additional supply to meet extraordinary demand for our Xeon family". Intel is selling server processors faster than it can make them, even as its largest market shrinks.
That combination explains the size of the reaction. Investors had bought Intel as an AI infrastructure supplier. The results confirmed the AI part of that story and undermined the PC part at the same time, so the multiple had to absorb both messages at once.
Why are PC shipments falling so fast?
Intel's PC problem is an industry problem. IDC's preliminary data, published on 8 October, put third-quarter PC shipments at 62.7 million units. That is down 20.1% from 78.5 million a year earlier, and 9.1% lower than the second quarter.
Gartner put the decline at 22.4% in its own release. The two firms use different definitions and vendor estimates. Both describe one of the steepest drops in years. IDC had already recorded a 3.8% fall in the second quarter.
The sequential fall breaks the normal pattern. Third quarters usually beat second quarters as retailers and businesses stock up for the holiday period. This year shipments fell 9.1% from the second quarter instead of rising.
The main cause is price. Memory and storage costs have climbed, and PC makers have passed the increases on to buyers. Gartner's release ties the shipment decline to component cost inflation that raised average selling prices.
IDC expects average selling prices to rise 17% in 2026. It does not expect meaningful relief from the memory shortage before the end of 2027. The firm forecasts an 11.3% fall in full-year shipments, with the fourth quarter down 20% year over year.
Demand was also pulled forward. Buyers stocked up in the first half of 2026 to beat expected price rises. IDC's Jean Philippe Bouchard said that anticipation led some end users "to make purchases earlier than anticipated", leaving channels with inventory just as prices peaked.
The vendor data shows how broad the damage was. Lenovo stayed the largest PC maker with 23.8% share despite a 22.6% decline. HP's shipments fell 30.9%, the steepest drop among the top five, and its share slipped to 16.5%. Dell held 12.1% after a 25.0% fall.
Apple declined less than the market, with shipments down 11.3%, so its share rose to 9.5% from 8.5%. ASUS shipped 5.5 million units, down 8.6%, and held 8.7% of the market. Smaller vendors together fell 14.0%.
- •Lenovo: 14.9 million units, down 22.6%
- •HP: 10.3 million units, down 30.9%
- •Dell: 7.6 million units, down 25.0%
- •Apple: 5.9 million units, down 11.3%
- •ASUS: 5.5 million units, down 8.6%
How much of Intel's business still depends on PCs?
PCs remain Intel's largest single revenue line. Client computing brought in $8.1 billion, or 49% of the $16.5 billion total, against $7.3 billion from data center and AI.
That balance is shifting quickly. Data center grew 78% while client fell 5%. A year ago the client unit was roughly twice the size of data center; now the two are close.
Intel argues that processor demand is structural rather than cyclical. Zinsner told analysts the company expects industry server CPU units to grow more than 30% in both 2026 and 2027, CNBC reported. Agentic AI systems use CPUs to coordinate work alongside graphics processors.
The applications matter here. Chatbots and large models lean heavily on GPUs from Nvidia and AMD. Agentic tools such as Meta's Muse and OpenAI's Dots use CPUs to orchestrate tasks and data across cloud machines, which is why server processor demand has held up while PC demand has not.
The difficulty is serving that demand. Tan's stated priority is extra Xeon supply, and the Israeli outage removes volume at an awkward moment. A company that is short of product in its growth business, and short of buyers in its core one, tends to be valued on its weakest segment. That is what the share price reflected.
The client unit also missed analyst estimates for the quarter, CNBC reported, even though group revenue was in line. Intel renamed the division in 2026, adding "Physical AI" to its title, but the revenue still rises and falls with PC shipments.
Scale still matters. Intel Products, the combined client and server business, generated $15.4 billion of the group total. Even after a year of rapid server growth, the PC franchise remains the base that pays for Intel's manufacturing expansion.
What is happening in Intel's foundry business?
Intel Foundry reported $6.7 billion of revenue, up 58%. Most of that is internal. Intersegment eliminations of $6.3 billion show that Intel's own product groups remain the foundry's main customers.
That accounting matters when reading the growth rate. A 58% increase in foundry revenue is not evidence of outside demand, because the wafers may simply be moving from one Intel unit to another. External customers are the number to watch, and Intel's release does not name any.
The segment is still losing money. 24/7 Wall St., citing Intel's SEC filing, reported that the foundry unit posted an operating loss in the quarter.
Manufacturing progress is real. Intel is in high-volume production of its 18A process at its new Arizona fab, CNBC reported, giving customers an alternative to Taiwan Semiconductor Manufacturing Co. for advanced logic. Intel also released a 0.9 version of its 14A process design kit in October and is working with ASML on readiness for High-NA EUV lithography.
The company is the only US-based manufacturer of advanced logic chips, according to CNBC. The US government took a 10% stake in Intel in 2025, and Nvidia and SoftBank have also invested billions.
Intel is also spreading its bets across packaging and specialty work. It completed the RAMP-C program, which enabled validated prototypes for secure enclave chips, and announced a packaging collaboration with Lens Technology around glass processing for advanced chip packaging.
Capital remains the risk. The equity offering strengthened the balance sheet, but the Ohio fab once expected to deliver chips in 2025 has only just had its first structural truss installed. Intel's next Arizona fab is due to begin operations by the end of 2028.
The strategic question is whether these plants fill with outside work. Intel's own products can absorb only so much capacity, and a foundry that mostly serves its parent is a cost center rather than the second growth engine investors expected.
What do the results mean for laptop buyers?
Cheaper laptops are unlikely in the next two quarters. IDC expects average selling prices to rise 17% in 2026 and does not see memory supply returning to normal before the end of 2027.
The product mix is shifting upward as well. IDC says manufacturers are struggling to maintain full product portfolios, which points to fewer low-cost configurations rather than discounts. When components are scarce, vendors put them in machines that earn more per unit.
The near-term demand picture is weak on both sides. IDC's June forecast expected fourth-quarter shipments to fall 20% year over year, and the third quarter already came in below the second. Waiting for a better price has recently meant paying more, not less.
Apple has been a partial exception. IDC said its new MacBook Neo drove stronger-than-expected notebook demand and that competitive pressure from it kept some low-cost notebook options alive. Even so, IDC expects average prices to keep climbing.
There is some new hardware. Intel said the first Googlebook devices built on its Core Ultra Series 3 processors will come from Acer, ASUS and Lenovo. Their pricing will be set in a component market that has moved against buyers all year.
For anyone replacing a laptop in the next six months, memory is the specification to watch. IDC expects prices to keep rising while capacity stays short, and it does not expect the market to return to 2025 pricing even after memory supply expands.
Business buyers face a different calculation. A fleet refresh can be delayed, but with prices still climbing, waiting does not guarantee a lower price. Buyers who must purchase now are negotiating in a market where vendors have little room to discount.
What to watch next for Intel and the PC market
AMD reports next week and Nvidia reports in mid-November, CNBC reported. Both will show whether the AI hardware boom is still accelerating or whether October's selloff reflected a wider reassessment of spending.
Intel's own fourth quarter is the real test. Guidance of $15 billion to $16 billion implies a sequential decline from $16.5 billion. Hitting even the top end would require data center growth to offset another weak PC quarter.
Three operational questions matter. The first is when the Israeli fab returns to full output. The second is whether memory prices ease in 2027, which decides how quickly PC demand recovers. The third is whether 18A and 14A win outside foundry customers rather than only internal ones.
Watch the language on PC demand as well. Intel said the slowdown runs through at least the first half of 2027. If that forecast shortens, the client business could stabilise sooner than the current guidance implies. If it lengthens, the company's growth will depend almost entirely on servers.
Memory contract prices are the other leading indicator. They set the pace of PC pricing, and IDC does not expect relief from the shortage before the end of 2027. Any early relief would show up in laptop prices before it shows up in Intel's revenue.
For readers, the simplest signal is the shipment data that IDC and Gartner publish each quarter. If those numbers keep falling by double digits, Intel's client business will keep dragging on a company that is otherwise growing quickly in servers.