Intel has announced a return to profitability, attributing this turnaround to increased revenue and a growing demand for artificial intelligence (AI) that may benefit its foundry services. On Thursday, the company released its Q3 FY 2025 results, noting a revenue of $13.7 billion—marking a three percent increase from the previous year—and a net income of $4.1 billion, a significant recovery from the $2.9 billion loss reported in Q2.
However, not all segments performed well; sales of Intel“s datacenter products declined by one percent, while its foundry division experienced a four percent revenue drop. The client products segment led the revenue growth with $8.5 billion, reflecting a five percent increase. CFO David Zinsner commented during the earnings call that a notable refresh of Windows has positively impacted desktop sales, as consumers are investing in new hardware to accommodate Windows 11 following the end of support for Windows 10.
Despite these gains, Intel faces challenges in the AI hardware market, where competitors like Nvidia and AMD dominate with their GPUs. The company”s foundry services are also lagging, struggling to compete with TSMC and still working to refine its 18A manufacturing process. During the earnings call, Intel executives highlighted their efforts in addressing these challenges. CEO Lip Bu Tan stated that significant progress has been made on the 18A process, particularly with the operationalization of Fab 52 in Arizona, which is set to produce silicon wafers for these advanced chips.
Nevertheless, Zinsner advised caution regarding the 18A ramp-up, indicating that Intel will not rush capacity increases without confirmed clients, emphasizing that the company aims to ensure that 18A is a reliable option for potential partners. Currently, only Microsoft has committed to utilizing this process.
Looking ahead, Intel is also engaging with chip design firms regarding its future 14A process, which is reportedly more developed than the 18A was at a similar stage. In terms of GPUs, Tan mentioned that the company plans to introduce successive generations of inference-optimized GPUs annually, with improvements in memory and bandwidth tailored to enterprise needs. This strategy acknowledges the anticipated rapid growth of hardware required for AI workloads, which increasingly demand inference capabilities.
Zinsner further suggested that Intel“s server CPU segment, historically a major revenue source, stands to gain from the expanding AI infrastructure. He expressed cautious optimism regarding the total addressable market (TAM) for server CPUs, projecting continued growth in 2026, despite acknowledging the need for competitive improvements. He noted that there is a growing interest among data center customers for long-term supply agreements to support their business objectives, driven by the swift development of AI infrastructure.
However, Intel cautioned that it still struggles to meet demand with its current production capabilities on the Intel 7 and Intel 10 processes and will not invest in increasing this capacity. Zinsner remarked that the company is partly relying on existing inventory to fulfill demand while also facing challenges in securing essential chipmaking supplies. Despite these hurdles, Intel has strengthened its financial position by reducing debt, which will provide the necessary capital for future investments. CEO Tan outlined ongoing hiring efforts aimed at bolstering the company”s chipmaking operations.
The market reacted positively to Intel“s announcements, resulting in a rise in share prices during after-hours trading, climbing from just over $38 to beyond $41.
