KIST Launches Korea's First AI Data Center Thermal Management Council

The Korea Institute of Science and Technology (KIST) launched Korea's first industry-academia-research council for AI data center thermal management on the 23rd. KIST held the '2026 AI Data Center Thermal Management Council Inauguration and Commemorative Symposium' at its Seoul headquarters in Seongbuk-gu. The council addresses heat generation challenges from high-performance semiconductors like GPUs and HBM as generative AI and large-scale AI services expand, requiring improved thermal management to enhance data center power efficiency and stability. Approximately 300 experts from over 45 institutions participated, including Samsung Electronics, SK Hynix, LG Electronics, Seoul National University, Sungkyunkwan University, Yonsei University, and the Electronics and Telecommunications Research Institute (ETRI).

Council Targets Standardization and Technology Roadmap Development

The council aims to standardize high-efficiency and high-reliability thermal management technologies for AI data centers, establish technology roadmaps, define testing and reliability standards, and promote joint demonstration projects. The initiative plans to enhance the competitiveness of Korea's AI data center industrial ecosystem by linking technologies across materials, devices, semiconductor packaging, and cooling systems.

Symposium Discusses Next-Generation Thermal Management Technologies

The symposium covered next-generation thermal management technologies including two-phase cooling, cold plates, immersion cooling, data center refrigerants, high thermal conductivity materials, active cooling, packaging reliability, and thermal property measurement for high-heat semiconductors.

FAQ

What did KIST launch on the 23rd?

KIST launched Korea's first industry-academia-research council for AI data center thermal management and held an inauguration symposium at its Seoul headquarters in Seongbuk-gu.

Why is thermal management technology important for AI data centers?

As generative AI and large-scale AI services expand, high-performance semiconductors like GPUs and HBM generate increased heat, making thermal management critical for improving data center power efficiency and stability.

Which organizations participated in the council?

Approximately 300 experts from over 45 institutions participated, including Samsung Electronics, SK Hynix, LG Electronics, Seoul National University, Sungkyunkwan University, Yonsei University, and ETRI.