Supercomputing

• On May 7, 2026, South Korea’s Ministry of Science and ICT announced the formation of a national AI supercomputing grid, integrating 16 existing data centers into a unified platform with total processing capacity of more than 40 exaFLOPS. • The grid, funded with ₩12 trillion (about $9 billion), will provide subsidized compute access to domestic startups, universities, and automakers, with particular emphasis on training Korean‑language LLMs and autonomous‑driving models. • Seoul positions the grid as a counterweight to China’s state‑backed AI clusters and U.S. hyperscalers, while also tightening export controls on high‑end GPUs integrated into the network.

• The U.S. Department of Energy's Genesis Mission, led by Dr. Dario Gil (formerly IBM senior vice president), is deploying AI supercomputing infrastructure to transform American science and engineering across fusion, energy, and national security. • Argonne National Laboratory will deploy approximately 10,000 state-of-the-art GPUs this year via Nvidia and Oracle, with a comparable AMD-HPE cluster launching at Oak Ridge National Laboratory in Tennessee, and a planned 100,000-GPU cluster at Argonne in 2027. • The mission aims to produce 50 to 100 comparable breakthroughs across all scientific domains within three to five years, creating a durable platform of AI supercomputers and next-generation quantum computers for the scientific community.