Intel has reported a notable financial rebound driven by a shift in artificial intelligence (AI) spending that now favors its central processing units (CPUs) alongside the graphics processing units (GPUs) traditionally dominated by competitors like Nvidia. The company’s latest quarterly results, released Thursday, showed a 25 percent increase in revenue to $16.1 billion, largely propelled by nearly 60 percent growth in its data center segment. This division supplies chips essential for AI applications and cloud computing services.

The earnings surpassed Wall Street expectations, prompting Intel’s stock to rise nearly 9 percent in after-hours trading. Intel Chief Executive Officer Lip-Bu Tan described the surge as the company’s strongest revenue growth in 15 years, attributing it to "unprecedented demand for compute" driven by advancements in AI.

Despite the positive sales momentum, Intel reported a significant net loss of $11 billion during the quarter, compared with a $2.9 billion loss a year earlier. This increased deficit primarily reflects restructuring expenses and a reassessment of shares related to an $8.9 billion investment made by the U.S. government last year. Nonetheless, Intel’s stock price has more than quadrupled since the start of the year, signaling investor confidence in the company’s recovery.

The shift in AI technology usage has played a key role in Intel’s resurgence. Earlier phases of the AI boom heavily favored GPUs, which excel at handling many simple operations simultaneously. CPUs had a supporting function but were typically outnumbered by GPUs by a factor of at least four to one in AI servers. However, the rise of AI “agents,” software capable of reasoning and complex decision-making, has increased the demand for CPUs, as these processors are better suited to the tasks these programs require. Industry analysts forecast that AI data centers will soon require roughly equal numbers of CPUs and GPUs.

This evolving landscape also benefits Intel’s competitor Advanced Micro Devices (AMD), which has expanded its presence in data center servers and developed GPUs tailored for AI workloads. On Thursday, AMD introduced new, more powerful CPUs, GPUs, and integrated “rack-scale” systems designed to consolidate processing, networking, and storage. The AI startup Anthropic has committed to purchasing AMD’s new hardware, with AMD forecasting the deal to be worth tens of billions of dollars and including up to a $5 billion investment from the chipmaker. OpenAI, Anthropic’s competitor, has similarly announced plans to utilize AMD’s rack-scale systems.

Meanwhile, Nvidia continues to develop its own AI-focused processors, including a new model named Vera, while major data center operators such as Amazon, Microsoft, and Google have created custom microprocessors tailored to their specific needs.

Intel maintains competitive advantages through its extensive network of manufacturing facilities, differentiating itself from foundry leaders like Taiwan Semiconductor Manufacturing Company (TSMC). Intel competes not only in chip fabrication but also in packaging services—a growing area where it has recently secured customers such as Apple, which had previously relied largely on TSMC.

Intel reported its Xeon 6 server chip among the fastest-selling products in its history amid record-breaking sales growth in its server chip business. Despite strong demand, the company said it had reduced some staffing in that segment to enhance operational efficiency, though it did not provide specific figures.