TAIPEI (Taiwan News) — National Yang Ming Chiao Tung University (NYCU) launched its first Nvidia-powered AI computing center on Tuesday (Aug. 6).

According to an NYCU press release on Tuesday, the platform includes two Nvidia H100 and three H200 servers, marking their first use in a Taiwanese university. The center will also feature 40 GPUs with a computational power of 1.7 PFLOPS, per Liberty Times.

At the inauguration ceremony, NYCU President Lin Chi-hung (林奇宏) said the center would help with AI research, per CNA. By combining efforts from alumni and industry, the school can enhance AI computational power, he said.

Department of Electrical Engineering Professor Chien Jen-Tzung (簡仁宗) said his research team focuses on AI model training. Experiments that took a full day can now be completed in just over two hours, significantly reducing experiment waiting times, he explained.

With a fourfold memory expansion, the school can train AI models with better performance and higher prediction accuracy, Chien added.

The school said over the next six months it will integrate three more H200 servers and incorporate an AI high-speed parallel file system into the platform.