TAIPEI, 21 November 2025 — Foxconn (TWSE:2317) Chairman Young Liu says the company now has the capacity to manufacture 1,000 artificial intelligence (AI) racks every week and expects that production rate to rise further next year, underscoring its rapid expansion into the AI infrastructure race.
Liu made the remarks at Foxconn’s annual tech day, attended by key partners including Nvidia (NASDAQ:NVDA) and OpenAI. Founder Terry Gou also made a rare appearance, signalling the event’s strategic importance.
A Major Beneficiary of the Global Data Centre Boom
Foxconn, best known as Apple’s (NASDAQ:AAPL) top iPhone assembler, has aggressively expanded beyond traditional electronics into AI data centres and electric vehicles (EVs).
It has now become Nvidia’s main manufacturer of AI racks, specialised server racks built for AI workloads containing GPUs, cables and high-speed components.
That positioning has made Foxconn a direct beneficiary of the global rush to build AI supercomputing capacity.
Liu previously told Reuters that Foxconn expects AI demand to be a major growth driver in 2026, and that the group plans to invest US$2–3 billion annually in AI.
Next Frontier: EVs, Quantum Computing, and Asia’s First GB300 Data Centre
Liu added that Foxconn’s EV production volumes have reached a “meaningful level”, enabling global automakers to increasingly consider Foxconn as an outsourcing partner. The company also plans to showcase new EV models designed specifically for Japan.
He said Foxconn was also working on quantum computing, reflecting its ambition to diversify into high-impact emerging technologies.
Separately, Neo Yao, CEO of Foxconn subsidiary Visionbay.ai, announced plans to deploy a 27-megawatt AI data centre in the first half of 2026.
The facility will house Nvidia’s most advanced GB300 chips, making it:
- Taiwan’s largest advanced GPU cluster, and
- Asia’s first GB300-powered AI data centre.
