Seoul National University of Science and Technology (President Lee Dong-hoon) announced on the 4th that it held a gap ceremony for its supercomputer data center.

A complete of KRW 4.09 billion was invested in the development of the supercomputer data center by the Ministry of Education and the university for the expansion of lab equipment, and it should be used as a core infrastructure to steer the education of the world’s best talent in the sector of ‘ChatGPT’, a super-giant AI. .

The GPU model ‘Nvidia DGX A100’ used for chat GPT learning was also applied to the supercomputer data center of Seoul National University of Science and Technology. It’s explained that it is feasible to implement an optimal AI computing cluster by demonstrating high performance in large-capacity AI learning and inference processing.

The computational performance of the supercomputer built by Seoul National University of Science and Technology reaches 40 petaflops in theory. That is the best level within the country.

Specifically, Seoul National University of Science and Technology’s supercomputer data center has applied ‘Uyuni’, an answer of CE Lab that maximizes GPU utilization. Uyuni creates a container with GPU resources allocated with an easy input, and constantly tracks the present task status and history.

Lee Dong-hoon, president of Seoul National University of Science and Technology, said, “Through the supercomputer infrastructure construction project, which is named the center of super-giant AI development, we’re contributing to fostering AI talents who will lead the 4th industrial revolution era.” “We are going to do our greatest to offer services that take note of the flexibleness and scalability.”

Reporter Juyoung Lee juyoung09@aitimes.com