Rittal
Megawatt cooling developed for AI
Rittal has developed a cooling solution that delivers over 1 MW of cooling power and paves the way for artificial intelligence. This is because the demand for computing power is growing so rapidly that a new level of scaling, cooling, power distribution and energy efficiency is required.
The power density for AI applications such as the training and operation of large language models (LLMs) in future data centers or the already widespread high performance computing will quickly push today's conventional air cooling to its physical and economic limits. Ultra-fast graphics processors (GPUs) produce so much heat that manufacturers are designing them for direct liquid cooling. In close cooperation with several hyperscalers, Rittal has developed a modular cooling solution that delivers a cooling capacity of over 1 MW with direct water cooling. Coolant Distribution Units are used for single-phase direct liquid cooling with water. For this purpose, the server in the rack ('Open Rack V3') is connected to the central inlets and outlets of the water circuit using standardized connections. Functional units such as the central controller unit and several coolant conveying units (CCUs), depending on the performance requirements, are completely modular and are simply pushed into the rack. Leakage monitoring starts at component level. When it comes to servicing, this concept offers the advantage that components such as controllers, sensors or the pump units of the in-row solution can be serviced during operation and replaced via 'hot swap'. Power is supplied via the standardized DC busbar in the rack.
The liquid-to-liquid solutions provide cooling as a bayed solution for racks over 1 MW, and in single racks up to 100 kW. Liquid-to-air variants are also available for data centers without a water connection, which dissipate the heat at the rear door of the rack or via a side cooler as a closed system to the air in the data center.










