U.S. based cloud services provider GMI Cloud announced on Monday that it will build a $500 million artificial intelligence data center in Taiwan. The project is backed by U.S. chipmaker Nvidia and is expected to come online by March 2026.
AI Infrastructure and Power
The new data center, which GMI Cloud CEO Alex Yeh referred to as a "strategic asset," will run on Nvidia’s new Blackwell GB300 chips. The facility will house approximately 7,000 GPUs
across 96 high density racks. It will be capable of processing nearly 2 million tokens per second and will draw around 16 megawatts of power.
Yeh emphasized that Taiwan needs more data centers to support its AI development, noting that the island's power supply challenges can be managed. He added that AI demand remains strong, with the company's existing GPU utilization being "almost full."
Industry Context and Business Impact
This deal is part of a massive global trend where technology giants are investing billions in AI infrastructure, creating a windfall for semiconductor companies like Nvidia. Nvidia CEO Jensen Huang often refers to such large scale processing centers as "AI factories."
GMI Cloud, which is one of Nvidia’s cloud partners and operates data centers in several countries, plans to seek an initial public offering (IPO) in the next two to three years. The Taiwan AI factory is expected to generate about $1 billion in total contract value once it is fully operational. Initial customers already secured for the facility include Nvidia itself, cybersecurity firm Trend Micro, electronics maker Wistron, and data infrastructure provider VAST Data. This project follows another major AI data center announcement in Taiwan made in May by Foxconn and Nvidia.
Found this article interesting? Follow us on X(Twitter) ,Threads and FaceBook to read more exclusive content we post.
