The expanded hardware aims to bridge the gap between traditional air-cooled data centers and the extreme thermal demands of modern AI factories. By offering up to 240kW of total cooling capacity at the rack level, these units allow operators to increase compute density without undergoing invasive facility overhauls. The systems are engineered for compatibility with standard EIA, ORv3, and MGX racks, providing a modular path toward more efficient energy management.
Supermicro Scales Liquid Cooling for High-Density AI Infrastructure
San Jose-based Supermicro is expanding its rack-level liquid cooling capabilities, launching ten new Rear Door Heat Exchanger models designed to support high-density AI and HPC clusters. The portfolio targets both modern facilities and legacy retrofits, offering cooling capacities that scale from 10kW up to 120kW per rack.
"We continue to expand our DCBBS offerings to provide our customers with unmatched customization and optimization options," said Charles Liang, president and CEO of Supermicro. The units can function as standalone cooling solutions or integrate alongside existing Direct-to-Chip technologies. Beyond hardware, the rollout includes intelligent management software, enabling real-time monitoring of flow rates, pump status, and temperature through platforms like SuperCloud Composer. This approach seeks to reduce integration risk and shorten the time required to bring new AI infrastructure online.
Comments (0)
No comments yet. Be the first!