Search results for SuperPod
AI & Enterprise
Govt GPU second-round bidding draws five contenders, Coupang re-entry in focus
South Korea\'s second round of government support for graphics processing units has attracted five bidders, including first-round winner Naver Cloud and first-round loser Coupang, both re-entering. KT Cloud, Samsung SDS and Elice Group joined as new participants. The science ministry set evaluation criteria including economic feasibility, large scale, direct clustering, latest GPUs and service launch within the year. Officials plan evaluations and site inspections in April and May before selecting operators in May.
AI & Enterprise
Naver team builds 4,000 Nvidia Blackwell B200 GPU cluster, says AI model development 12 times faster
Naver said on Jan. 8 it has completed an AI computing cluster based on 4,000 Nvidia B200 Blackwell GPUs. The company said the new “B200 4K cluster” is comparable in computing scale to leading TOP500 supercomputers. Internal simulations showed training a 72 billion-parameter model could be cut from about 18 months on an A100-based 2,048-GPU setup to about 1.5 months, improving training efficiency by more than 12 times.
AI & Enterprise
Nvidia unveils DGX SuperPOD supercomputer for Rubin chips
Nvidia unveiled the DGX SuperPOD, a Rubin platform-based supercomputer, at CES 2026 in Las Vegas on Jan. 7. The company said it presented a blueprint for AI infrastructure aimed at agentic AI and large-scale inference. DGX SuperPOD integrates multiple DGX Vera Rubin NVL72 systems with 576 Rubin GPUs, delivering 28.8 exaflops of FP4 performance and 600 terabytes of high-speed memory. Nvidia also introduced the Rubin platform, built on six new chips.