Artificial intelligence (AI) infrastructure Elice said on the 16th that it was selected as a supplier for the "2026 national AI data center advancement project" led by the Gwangju Artificial Intelligence Industry Convergence Project Group (AICA).
The project led by AICA provides high-performance AI computing resources and services specialized for research, development, and service to domestic corporations, institutions, and universities. Participating as one of several suppliers in the project, Elice will supply a high-performance graphics processing unit (GPU) cloud infrastructure through "Elice Cloud." The agreement period is about seven months, from May this year to the end of December.
Elice supports a high-performance GPU cloud infrastructure consisting of Nvidia H100 and the latest GPU B200. It also provides InfiniBand-based high-performance computing (HPC) multinode and single-node environments required for large-scale computation, as well as Cloud Security Certification (CSAP) zone resources for responding to the public institutional sector.
The company's optimization technology for its self-developed modular AI data center, "Elice AI PMDC," will also be applied. Elice AI PMDC features high-density power infrastructure and a high-efficiency cooling system. It is specialized for stably running high-performance GPU accelerators.
Based on this infrastructure technology, Elice plans to provide integrated support for overall AI cloud infrastructure operations, including incident response under a service-level agreement (SLA) and security monitoring.
Chief Executive Kim Jae-won of Elice said, "Through this project, Elice will provide a proven GPU cloud infrastructure to reduce constraints in research and development environments, and, based on our experience operating large-scale infrastructure, contribute to advancing Korea's AI infrastructure and strengthening national AI competitiveness."