Intel delivered a stronger-than-expected second quarter and issued an upbeat revenue forecast for the current period, reinforcing investor expectations that the semiconductor company is beginning to benefit more directly from the global expansion of artificial-intelligence infrastructure.

The Santa Clara, California-based chipmaker reported revenue of $16.13 billion for the quarter ended June 27, an increase of 25% from $12.86 billion in the same period of 2025. The result surpassed the $13.8 billion to $14.8 billion range Intel had provided three months earlier and represented the company’s fastest annual revenue growth in more than 15 years, according to management.

Intel forecast third-quarter revenue of between $15.8 billion and $16.8 billion. The midpoint of $16.3 billion would represent approximately 19% growth from the comparable quarter a year earlier and stood above the roughly $15.1 billion consensus estimate cited by MarketWatch before the announcement. Intel shares advanced sharply in after-hours trading as investors responded to the revenue outlook and stronger adjusted profitability.

The company expects third-quarter GAAP diluted earnings of approximately $0.31 per share and non-GAAP earnings of $0.38 per share. It projected a GAAP gross margin of 41% and a non-GAAP margin of 42%, with both the margin and earnings forecasts calculated at the midpoint of the revenue range.

Chief Executive Lip-Bu Tan attributed the quarter’s growth to accelerating demand for computing capacity across Intel’s processor, custom-chip, packaging and manufacturing businesses. He said artificial intelligence was generating unprecedented demand for compute and described the quarter as evidence that efforts to increase organizational speed, accountability and customer focus were improving Intel’s execution.

Chief Financial Officer Dave Zinsner said Intel exceeded its prior financial guidance because of robust demand, higher factory yields and improved manufacturing cycle times. Those improvements allowed the company to ship more products than anticipated and helped lift profitability despite continuing supply constraints across parts of the semiconductor industry.

Intel’s Data Center and AI division produced the fastest growth among the company’s major product businesses. Segment revenue rose 59% year over year to $6.3 billion as cloud-service providers, enterprises and AI infrastructure operators increased purchases of server processors and related technologies.

The result illustrates the continuing importance of CPUs within AI systems. Although graphics processors and dedicated accelerators perform much of the intensive model-training and inference work, servers still require general-purpose processors to manage data movement, storage, security, networking and orchestration. Intel is seeking to position its Xeon portfolio as a core component of increasingly complex AI clusters rather than competing only in the accelerator market.

During the quarter, Intel introduced Xeon 6+, its first server-class product manufactured with the Intel 18A process. The company said the processor was designed to sustain performance under real-world power constraints, an increasingly important issue for data centers facing limits on electricity availability, cooling systems and physical space.

Intel also expanded collaborations around rack-scale AI infrastructure and disaggregated inference. It highlighted work with SambaNova and Foxconn on production-ready systems for inference and agentic AI workloads, as well as a Vector Core Compute architecture combining Intel Xeon processors, SambaNova reconfigurable dataflow units and Nvidia Blackwell graphics processors.

The strategy reflects Intel’s attempt to participate in AI deployments even when customers select accelerators from competing suppliers. By providing host processors, networking products, custom silicon, packaging and manufacturing services, Intel can potentially earn revenue across multiple layers of an AI system rather than depending on a single accelerator platform.

Revenue in Intel’s Client Computing and Physical AI Group increased 13% to $8.9 billion. The renamed segment includes the company’s established personal-computer processor business as well as products intended for robotics, industrial systems and other edge-computing applications.

The client result suggests that Intel continued to benefit from commercial PC replacement activity and growing demand for devices capable of running AI workloads locally. The company said more than 130 customers were adopting or testing its Core Ultra Series 3 and Core Series 3 processors for edge AI and robotics applications.

Intel also introduced OpenVINO Physical AI, an open-source framework intended to help developers deploy robotics models involving vision, language, reasoning and motion control. The initiative supports Intel’s effort to extend its processor franchise beyond conventional laptops and desktops into autonomous machines, industrial equipment and embedded systems.

Total Intel Products revenue reached $15.1 billion, an increase of 28% from the prior-year period. The growth in product revenue was particularly significant because Intel has simultaneously been restructuring its manufacturing operations and reducing expenses following several years of market-share losses, process delays and weak profitability.

Intel Foundry reported revenue of $5.8 billion, up 31% year over year. The figure includes manufacturing services provided to Intel’s own product divisions and therefore should not be interpreted as equivalent to external foundry sales. Intel recorded $5.5 billion of intersegment eliminations during the quarter, reflecting the substantial internal component of the foundry segment’s reported revenue.

Nevertheless, the improvement indicates that higher internal production volumes and better factory performance are beginning to support the economics of Intel’s manufacturing network. The company said Intel 18A-P entered risk production on schedule, while a subset of Core Ultra Series 3 processors, code-named Panther Lake, entered high-volume manufacturing using ASML’s High-NA extreme-ultraviolet lithography technology.

Intel’s ability to improve yields on Intel 18A and related processes remains central to its recovery. Higher yields increase the number of usable chips obtained from each wafer, lowering unit costs and making additional supply available for customers. Faster cycle times also reduce the period between the start of wafer production and finished-product delivery, allowing Intel to respond more effectively to changes in demand.

The company is seeking external customers for its foundry operation while continuing to manufacture a larger share of its own most advanced products internally. Intel has described advanced packaging, purpose-built silicon and geographic supply-chain diversification as potential advantages in attracting customers that want alternatives to Asian manufacturing capacity.

During the quarter, Intel announced a collaboration with cybersecurity company Fortinet to develop a new security processor using Intel’s design, packaging and manufacturing capabilities. It also outlined a €5 billion investment to expand production capacity for current and next-generation Xeon processors manufactured on the Intel 3 process.

The stronger revenue performance translated into a substantial improvement in operating results. Intel’s GAAP gross margin increased to 40.4% from 27.5% a year earlier, while non-GAAP gross margin rose to 41.8% from 29.7%. The company recorded GAAP operating income of $1.8 billion, equivalent to an 11.1% operating margin, compared with a $3.18 billion operating loss and a negative 24.7% margin in the prior-year quarter.

On a non-GAAP basis, operating margin improved to 17.2% from negative 3.9%. Research and development plus marketing, general and administrative expenses declined 6% on a GAAP basis to approximately $4.5 billion. Non-GAAP operating expenses in those categories fell 8% to $4 billion.

The combination of higher revenue, lower expenses and improved manufacturing efficiency produced non-GAAP net income attributable to Intel of $2.2 billion, or $0.42 per diluted share. That compared with a non-GAAP loss of $441 million, or $0.10 per share, in the second quarter of 2025. Intel’s April outlook had called for non-GAAP earnings of approximately $0.20 per share.

The GAAP result was significantly weaker because of a large non-operating accounting adjustment. Intel reported a GAAP net loss attributable to the company of $11.03 billion, or $2.16 per share, compared with a loss of $2.92 billion, or $0.67 per share, a year earlier.

The primary difference between the GAAP and adjusted results was a $12.53 billion mark-to-market loss on escrowed shares. Intel excluded that charge from its non-GAAP calculation, along with share-based compensation, acquisition-related adjustments and restructuring expenses. The accounting loss did not erase the underlying improvement in the company’s operating income, but it created a pronounced gap between the two profit measures.

Intel generated $7 billion in cash from operations during the quarter, up from about $2.05 billion a year earlier. Gross capital expenditures declined to $2.65 billion from $4.49 billion, although management indicated that investment would rise as the company responds to stronger demand and expands manufacturing capacity.

Adjusted free cash flow was negative $8.42 billion, largely reflecting a $12.22 billion net outflow classified as partner contributions. The quarter followed Intel’s repurchase of a partner’s interest in an Irish wafer-fabrication joint venture, a transaction that increased the company’s ownership of manufacturing assets but required substantial financing and cash resources.

Intel ended the quarter with $12.87 billion in cash and cash equivalents and $16.85 billion in short-term investments. Total debt, including short-term borrowings, was approximately $50.54 billion, up from about $46.59 billion at the end of 2025. Inventory increased to $12.49 billion from $11.62 billion.

The balance-sheet figures highlight the financial demands of Intel’s dual strategy: rebuilding a competitive product portfolio while creating a foundry business capable of serving both internal and external customers. Semiconductor factories require years of investment before reaching high utilization, while each new process generation carries substantial technical and execution risk.

Zinsner said the company would meaningfully increase investment in equipment, clean-room space and substrates to support anticipated growth during 2026 and 2027. Intel nevertheless maintained an emphasis on cost control, projecting approximately $23 billion of GAAP operating expenses and $16.5 billion of non-GAAP operating expenses for the full year.

Another factor affecting year-over-year comparisons was the deconsolidation of Altera. Intel sold a 51% stake in the programmable-chip business in September 2025, after which Altera’s operating results were no longer fully included in Intel’s consolidated statements. Intel cautioned that this structural change affected the comparability of reported figures between the two periods.

The company’s workforce also remained substantially smaller than a year earlier following restructuring and divestitures. Intel reported approximately 77,600 employees within its core operations at the end of the quarter, compared with 96,400 a year earlier. Including Mobileye and other subsidiaries, total employment was approximately 82,300, down from 101,400.

The third-quarter guidance now shifts attention toward the sustainability of Intel’s revenue acceleration. Investors will be watching whether server demand remains strong, whether PC growth persists despite industry memory and substrate constraints, and whether manufacturing yields continue improving as newer products move into volume production.

They will also scrutinize gross margin. Intel’s projected non-GAAP margin of 42% would represent another sequential improvement, but it remains below the levels the company historically generated before its manufacturing setbacks. Maintaining margin expansion while increasing capital investment will be important to demonstrating that higher revenue can produce durable cash returns.

The foundry business presents a longer-term test. Internal manufacturing volumes can improve factory utilization, but the strategic case for Intel Foundry ultimately depends on securing meaningful external commitments, delivering process technology on schedule and earning acceptable returns from facilities that require tens of billions of dollars in investment.

For the immediate earnings cycle, however, the report showed a company with stronger demand, greater available supply and improving operating leverage. Revenue exceeded the top of Intel’s prior forecast by more than $1 billion, adjusted earnings more than doubled the company’s earlier projection, and the third-quarter outlook surpassed market expectations.

Those results do not eliminate Intel’s financial, competitive or execution risks. The company remains behind leading accelerator suppliers in major portions of the AI market, carries substantial debt and must fund an ambitious manufacturing roadmap. Yet the second quarter strengthened the argument that CPUs, advanced packaging and domestic semiconductor production can give Intel several paths to participate in the AI infrastructure expansion.