Particle.news

Foxconn and Intel Announce Full-Stack Partnership to Build Rack-Scale AI Infrastructure

The pact aims to tackle data-center bottlenecks from larger AI models by pairing Intel’s processors and software with Foxconn’s manufacturing and systems-integration scale.

Overview

  • The partnership, announced Thursday at Foxconn’s annual tech day in Taipei, covers development from silicon and chips up to integrated server racks, modules, systems and applications.
  • Companies said they will work on Intel server racks, high-speed interconnects, system telemetry and advanced cooling and are exploring the option of custom AI chips.
  • Foxconn also disclosed an expanded collaboration with South Korea’s SK Group on large AI data centers and next-generation memory, showing a broader push into infrastructure beyond device assembly.
  • Neither company released product details, costs or launch timelines, and markets reacted quickly with Intel shares rising and Foxconn shares falling after the announcements.
  • The move positions Foxconn and Intel to compete with dominant AI-accelerator suppliers and could give cloud providers and industrial users more options for dense AI deployments and edge or robotics use cases.