By Annika Masrani
Publication Date: 2025-11-17 12:31:00
GMI Cloud is stepping deeper into the AI infrastructure boom. The U.S.-based GPU-as-a-Service provider said Monday it will build a $500 million artificial intelligence data center in Taiwan, a project that will run on Nvidia’s (NVDA) new Blackwell GB300 chips and come online by March 2026. Bit by bit, Taiwan is becoming a major hub for next-generation compute, even as the island continues to wrestle with power-supply constraints.
Meet Your ETF AI Analyst
The new facility will house about 7,000 GPUs across 96 high-density racks and support nearly 2 million tokens per second of compute. Power draw is expected to land around 16 megawatts. For an AI cluster of this scale, these numbers point to a serious step up in capability tied directly to rising demand from global enterprises and model developers.
GMI Cloud founder and CEO Alex Yeh said Taiwan needs more of these facilities because they act as “strategic assets” for local AI development. He added that power issues can be solved and said GPU utilisation at the company is already “almost full,” which helps explain why expansion is arriving so aggressively. Yeh said, “You want to promote local ecosystems, you have to build the data center first, you have to build the AI cluster first.”
Nvidia Extends Global Footprint Through New AI “Factories”
The project gives Nvidia yet another high-profile cluster win as tech giants compete to lock in compute capacity. Nvidia CEO Jensen Huang has often…
