2 hours ago
453
Foxconn said on Friday that it has started building a $1.4 billion AI supercomputing center with Nvidia, and the company expects the site to be ready in the first half of 2026.
When the doors open, Foxconn said the facility will be Taiwan’s largest advanced GPU cluster, built for heavy AI workloads that need steady power, low‑latency systems, and enough compute to feed nonstop model training.
The company added that the center will run 27 megawatts and use Nvidia’s Blackwell GB300 chips. Neo Yao, who runs Foxconn’s new AI unit called Visonbay.ai, said the site will become Asia’s first GB300 data center.
At Foxconn’s tech day, Nvidia vice president Alexis Bjorlin told the room that the cost of building separate compute sites keeps rising and said very directly, “As GPU technology accelerates, building individual facilities may no longer make economic sense.”
She added that “Renting compute resources may offer a far better return on investment, enabling flexibility and enabling companies to scale their compute according to both product and business cycles.”
The event was attended by partners including Nvidia, OpenAI, and Uber. Yeah, the guest list alone tells you the scale of what Foxconn is chasing here.
Foxconn expands as Nvidia partnership deepens
Foxconn is best known for assembling iPhones, but Young Liu said in a Reuters interview earlier on Friday that the company is pushing deeper into AI infrastructure and electric vehicles.
He said Foxconn now builds Nvidia’s AI racks, which are full rack systems with chips, cables, and cooling built for AI tasks.
Because of that role, Foxconn has become one of the biggest winners of the global data center boom as cloud companies spend billions on AI upgrades. Liu said demand linked to Nvidia will be a major driver of 2026 growth.
Liu also said Foxconn will invest $2 billion to $3 billion every year into AI. The company already has the capacity to build 1,000 AI racks per week, and he said that number will rise in 2025.
Foxconn’s founder Terry Gou showed up at the event, as did Spencer Huang, a product manager in Nvidia’s robotics group and the son of Nvidia founder Jensen Huang. Spencer Huang said Nvidia is working with Foxconn to move more AI tools into factories and production lines.
Liu added that Foxconn’s EV output is now at the point where automakers can start sending more work to the company. During the event, Chief Strategy Officer Jun Seki presented Foxconn’s Model A EV.
Liu said the car was designed by Japanese engineers and that Foxconn plans to set up a company in Japan to serve clients there. He said the Model A will also be built in Japan once production ramps up.
Join a premium crypto trading community free for 30 days - normally $100/mo.
English (US)