Nvidia and Foxconn announced plans to build Taiwan’s largest supercomputer, set to be housed at the Hon Hai Kaohsiung Super Computing Center.

The system will use Nvidia’s forthcoming Blackwell architecture, specifically the company’s GB200 NVL72 data center platform.

Nvidia_wk6Y1B5.width-880
– Sebastian Moss

Optimized for AI and high-performance computing, the GB200 NVL72 platform includes 64 racks and 4,608 Tensor Core GPUs. Each rack features 36 Nvidia Grace CPUs and 72 Nvidia Blackwell GPUs, connected via Nvidia NVLink technology, delivering 130 TBps of bandwidth.

According to the two companies, the supercomputer is slated to have more than 90 exaflops of AI performance, putting it on par with Germany’s Jupiter supercomputer and making it the fastest system in Taiwan.

Construction of the new unnamed machine has already begun, with the first phase due to be operational by mid-2025 and full deployment expected in 2026. Once operational, the supercomputer will be used to power cancer research and large language model development, whilst also integrating with robotics platforms and digital twin technologies.

“Powered by Nvidia’s Blackwell platform, Foxconn’s new AI supercomputer is one of the most powerful in the world, representing a significant leap forward in AI computing and efficiency,” said Foxconn VP and spokesperson James Wu.

The announcement was made on the same day that Foxconn’s chair, Young Liu told customers the company is planning to build a plant in the Mexican city of Guadalajara to build Nvidia GB200 Blackwell AI servers.

First reported by the FT, few details about the factory have been revealed however, the news outlet stated that at 450m-long, the facility would be the largest in the world dedicated to building Nvidia servers.