AI CPU Demand Sparks Growth: Intel's Q1 Revenue Beats Expectations, Stock Soars 16% After Hours
Intel released its Q1 earnings report on Thursday after the US stock market closed, with performance significantly exceeding Wall Street expectations. The struggling chip giant is finally showing signs of recovery, with its stock surging 16% in after-hours trading.
Here's a comparison of the company's actual performance and analysts' expectations from Refinitiv:
Adjusted Earnings Per Share: $0.29, compared to an expected $0.01.
Revenue: $13.8 billion, compared to an expected $12.42 billion.
Intel has recently become a Wall Street favorite, with its stock up over 80% year-to-date as of Thursday's close, and a full-year increase of 84% in 2025. The Trump administration strongly supported Intel, becoming its largest shareholder after a government investment last year to promote the repatriation of the chip industry to the United States. Nvidia and SoftBank have also invested billions of dollars in Intel.
Previously, in the early stages of the artificial intelligence industry, Intel's technology lagged far behind Nvidia and AMD, and its long-term development momentum was weak.
However, this situation is now turning around: the company's revenue increased 7.2% year-on-year, from $12.67 billion in the same period last year to $13.8 billion. Prior to this, Intel had experienced year-on-year revenue declines in five out of the previous seven quarters.
Intel's Q2 performance guidance: revenue between $13.8 billion and $14.8 billion, and adjusted earnings per share of $0.20, significantly higher than analysts' expectations of $13.07 billion in revenue and $0.09 per share.
The data center business was the main driver of this growth. Driven by a surge in CPU demand, Intel is gradually opening up its business in the field of artificial intelligence, with revenue in this sector increasing by 22% year-on-year to $5.1 billion.
The CPU market, which had been dormant for a long time, has experienced an explosion, and the demand for computing power related to intelligent agents is no longer solely reliant on Nvidia GPUs, which have long monopolized the AI field. The increase in CPU demand has also supported Intel's recent large-scale investments: spending $14 billion to repurchase 49% of the shares of its Irish wafer factory previously sold to Apollo Global Management.
Intel is still operating at a loss: the company's net loss widened to $4.28 billion, or $0.73 per share, compared to a net loss of $887 million, or $0.19 per share, in the same period last year.
Intel adopts a rare business model in the industry: as an Integrated Device Manufacturer (IDM), Intel both independently researches and produces end-chip products and develops its own chip manufacturing processes. Most chip companies outsource the complex and high-cost manufacturing process to large wafer foundries such as TSMC.
Intel's foundry business revenue increased 16% year-on-year to $5.4 billion, but most of this capacity is still used to produce its own chips.
Intel's 3rd generation Core Ultra processors have been available for installation since January of this year, and the new Xeon 6 + data center processors were launched in March. Subsequently, Google announced that it will use multiple generations of Intel CPUs for its data center AI computing business.
Intel's latest PCs and data center processors are both produced using the 18A process at its newly built large wafer factory in Arizona. Although this process technology is comparable to TSMC's 2nm process, Intel is currently the only major customer of its own 18A wafer factory.
Intel's future challenge lies in convincing customers who have long used TSMC to switch to its foundry services.
Several of Intel's process iterations have been delayed multiple times, and the 18A wafers have yield issues, meaning the number of usable chips produced per wafer is low. Some analysts are still watching the yield performance of Intel's next-generation 14A process, which is planned to be implemented in 2028 and beyond. Intel has stated that it will invest in mass production of new processes after finalizing cooperation with major customers; and CEO Pat Gelsinger said on social media in January that Intel will fully deploy the 14A process.
The core major customer that the market has been waiting for is likely Elon Musk, but details have not yet been fully disclosed. Intel announced earlier this month that it will move into Musk's Terra Foundry Industrial Park in Austin, Texas, to extensively research, develop, and manufacture high-performance chips for SpaceX, AI company xAI, and Tesla.
During Tesla's Q1 earnings conference call, Musk revealed that Tesla plans to use Intel's future 14A process to produce chips in the industrial park for Tesla's entire vehicles, humanoid robots, and SpaceX's planned orbital data centers.
Musk said that Intel's 14A process is still in development, but "once the Terra Foundry Industrial Park is mass-produced, the 14A process will be mature enough for commercial use."
Intel's strategic restart of its foundry business began during Pat Gelsinger's tenure as CEO in 2021. Gelsinger stepped down in 2024, and Chen Lifu took over as CEO at the beginning of last year.
In the same year, Intel cut 15% of its workforce and canceled construction projects for wafer factories in Germany and Poland. The launch of the large wafer factory project in Ohio was also postponed from this year to 2030. Chen Lifu wrote in a layoff memo that the company's investment scale in recent years has been too large and too fast, far exceeding actual market demand.