Intel has spent years trying to prove it still belongs at the center of the semiconductor industry. Its latest quarterly results offer something the company badly needed: numbers that do not require much explaining.
Strong demand for processors used in artificial intelligence servers pushed Intel’s second-quarter revenue to $16.1 billion, up 25% from the same period last year. The company described it as its fastest quarterly revenue growth in more than 15 years.
Now Intel expects that momentum to continue.
For the third quarter of 2026, the chipmaker forecast revenue between $15.8 billion and $16.8 billion. It also expects adjusted earnings of $0.38 per share, giving investors a more confident outlook than Intel has been able to offer during much of its recent turnaround.
AI Demand Is Giving Intel’s Server Business New Life
The loudest number in Intel’s report came from its Data Center and AI division.
Revenue in that business climbed 59% year over year to $6.3 billion. That is a sharp jump for a company often discussed as an AI laggard beside Nvidia and, increasingly, AMD.
Intel is not suddenly replacing Nvidia’s graphics processors inside the largest AI training clusters. That is not really the story here.
AI data centers need more than accelerators. They also require central processing units to manage workloads, move data, run cloud services and handle the less glamorous parts of operating enormous computing systems. As companies build more AI infrastructure, demand for those server CPUs grows alongside demand for GPUs.
Intel appears to be catching that secondary wave.
The company said AI-driven computing demand continues to strengthen across its CPU business, custom chips, advanced packaging operations and manufacturing network. Chief Executive Lip-Bu Tan called the current demand for computing capacity unprecedented.
Big language. This quarter, at least, the sales figures support it.
Intel Beat Its Own Forecast by a Wide Margin
Intel entered the second quarter expecting revenue of between $13.8 billion and $14.8 billion. It finished with $16.1 billion.
That is not a small beat caused by accounting noise or a slightly better product mix. Intel shipped considerably more hardware than it had expected.
Chief Financial Officer Dave Zinsner credited stronger customer demand, higher factory yields and improved manufacturing cycle times. In other words, more buyers wanted Intel chips, and the company became better at getting those chips out of its factories.
Adjusted earnings reached $0.42 per share, compared with an adjusted loss of $0.10 per share a year earlier. Intel also generated $7 billion in cash from operations during the quarter.
The official accounting result looked much uglier. Intel reported a GAAP net loss of roughly $11 billion, largely reflecting a mark-to-market adjustment connected to escrowed shares. Its adjusted net income, which excludes that and several other items, came to $2.2 billion.
That gap matters. Investors will probably spend less time looking at the headline loss and more time asking whether the underlying improvement can last.
The PC Business Also Grew
AI servers grabbed the attention, but Intel’s client computing business did not stand still.
Revenue from the Client Computing and Physical AI Group reached $8.9 billion, up 13% from the previous year. The division includes processors used in laptops, desktops and other personal computing devices.
The PC market has not returned to its pandemic-era frenzy. Still, businesses continue replacing ageing machines, while manufacturers are pushing a new generation of computers marketed around local AI features.
Intel wants to be inside those systems.
The company has already tied its PC strategy closely to processors with built-in neural processing capabilities. Whether consumers genuinely need an “AI PC” remains debatable, but device makers are moving in that direction anyway. Intel benefits as long as the replacement cycle keeps moving.
Intel Foundry Is Growing, Though the Bigger Test Remains
Intel’s manufacturing business reported revenue of $5.8 billion, a 31% increase from the same quarter last year. Most of that revenue still comes from producing Intel’s own chips, so it should not be viewed in the same way as outside customer sales at a contract manufacturer such as TSMC.
Even so, the growth shows that more production is flowing through Intel’s factory network.
The company is increasing investment in manufacturing equipment, clean-room space and chip substrates to support expected demand across both its product and foundry operations.
That introduces a familiar risk.
Building advanced semiconductor factories consumes enormous amounts of money. Intel needs enough customer demand to keep those facilities busy, and it still has to persuade major outside chip designers that its manufacturing processes can deliver reliably at scale.
One strong quarter does not settle that argument.
CPUs Still Matter in the AI Race
Much of the AI chip conversation revolves around GPUs and dedicated accelerators. Understandably so. Those components perform the heavy mathematical work behind training and running advanced AI models.
Yet a data center cannot operate on accelerators alone.
CPUs coordinate systems, run databases, manage networking tasks and support software that sits around AI workloads. The rapid construction of AI infrastructure is therefore expanding the total market for server components, not only the market for Nvidia-style accelerators.
That dynamic gives Intel room to grow without directly defeating Nvidia in the most advanced AI training market.
Intel is also promoting rack-scale infrastructure and disaggregated inference systems built around its Xeon processors. Recent demonstrations with SambaNova, Foxconn and Vector Core Compute combined Intel CPUs with other companies’ AI accelerators, including Nvidia Blackwell GPUs.
It is a practical approach. Intel does not need every chip in the rack. It needs its processors to remain difficult to remove.
Intel’s Turnaround Finally Has Better Numbers Behind It
Intel still faces uncomfortable questions about manufacturing costs, outside foundry customers and its ability to compete in specialised AI accelerators.
Those problems did not disappear in three months.
What changed is the tone of the evidence. Revenue grew 25%. Data center sales jumped 59%. Gross margin improved sharply, and the company’s third-quarter forecast came in well above the level many analysts expected.
Intel has talked about rebuilding for years. This quarter looked less like another promise and more like an actual step.
The AI boom may not be turning Intel into the next Nvidia. It could still give the company something nearly as valuable: a much larger market for the server processors, packaging services and factories it already knows how to build.
Sources
- Reuters – Intel forecasts upbeat quarterly revenue and profit on strong AI-driven server chip demand
- Intel – Second-Quarter 2026 Financial Results
- Intel – First-Quarter 2026 Financial Results