Seoul, South Korea — Intel said global demand for data-center CPUs is set to grow fivefold over the next few years, surpassing $200 billion, driven by the shift from generative AI to agentic AI. The forecast was presented on Thursday (12) by Anil Nanduri, vice president of Intel's Data Center Group, at a press briefing during the Intel AI Summit Seoul 2026 at COEX in Gangnam-gu, Seoul.
According to Nanduri, "as AI inference evolves into agentic AI, data centers are no longer just a matter of GPUs." With more AI models running, more tasks to process and more calls to external programs, CPUs are taking on a central role again — and as interaction shifts from human-AI to machine-to-machine, the roles of memory and networks grow as well.
Tight supply and a Korean alliance
Bae Tae-won, president of Intel Korea, said "significant difficulties in CPU supply are also expected next year" and that the company is consulting headquarters on how to prioritize demand and supply together with the government, including the server makers taking part in the K-AI Full Stack Alliance, officially launched on Thursday.
The alliance brings together 23 companies — including FuriosaAI, DeepX, Upstage, MegazoneCloud and Hancom — to build open AI infrastructure based on South Korean-made servers, focused on sovereign AI and physical AI. It spans hardware, software and services across five areas: sovereign AI, heterogeneous AI infrastructure, custom ASICs, edge and physical AI, and agentic AI and AI PCs, plus the startup ecosystem.
Partnerships in health, manufacturing and robotics
Dan Rodriguez, vice president of Intel's Edge Computing Group, announced cooperation with Samsung Medison to automate an ultrasound diagnostic platform using the built-in AI features of the Intel Core Ultra series, and with Daim Technologies on a reference architecture for autonomous factories integrating robots and industrial machinery.
Rodriguez added that Intel works on edge and physical AI not only with major Korean conglomerates but also with startups, and Bae said there are "extensive collaborations" with domestic robotics companies, without naming them. The company also cited Compute Express Link (CXL) as an area of collaboration with South Korea's semiconductor industry, and said its Open Edge ecosystem now includes more than 180 qualified AI systems from over 60 hardware partners and more than 160 independent software vendors.
Context: days after Nvidia in Seoul
The Intel AI Summit Seoul 2026 comes days after the NVIDIA AI Day Seoul 2026, held in the same city on November 9, when Nvidia gathered Samsung, SK and LG around physical and sovereign AI. Intel, for its part, said it does not compete over announcements: "what is important is working substantively with Korean partners and how to scale this up," Bae said. See Geekiki's coverage of Nvidia's event: NVIDIA AI Day Seoul 2026 começa com foco em IA física e agêntica, Nvidia apresenta Agentic AI Stack no AI Day Seoul and NVIDIA AI Day Seoul 2026 reúne Samsung, SK e LG.
Sources
- MoneyToday — Intel says AI boom has driven fivefold surge in CPU demand, expands ties with Korean server, robotics and medical firms
- AJU Press — Intel recites AI mantra in Seoul amid Nvidia turmoil
- BigGo Finance — Intel Launches 'K-AI Full-Stack Alliance' with 23 South Korean Companies
Featured image: Intel Xeon Scalable processor. Intel press photo, via TecnoGaming.
Transparency: this article was produced with the help of artificial intelligence, based on cross-checking public sources on the internet and public posts on X, with editorial review. The projections cited are statements from Intel, not established facts. Always consult the original sources listed above.
By Muse
