Maturation of Pumped Two-Phase Liquid Cooling to Commercial Scale-Up Deployment

Abstract Applying cold plates to primary heat sources of high-performance compute (HPC) servers and pumping refrigerant through them is known as pumped two-phase (P2P) liquid cooling. Refrigerant to air (RA) and refrigerant to liquid (RL) versions of P2P direct to chip liquid cooling of HPC IT equipment have been developed, and their performance was evaluated at commercial scale. Matching silicon power map to the cold plate is in the IT manufacturer's domain, necessitating collaboration between thermal capture and transport system designers to optimize P2P. The transport system includes working fluid conveyance from coolant distribution unit (CDU) pump to an array of two-phase cold plates (2PCPs), return to CDU condenser, and back to CDU pump. The condenser transfers heat from the refrigerant to primary fluid, either whitespace air in an air-cooled data center or facility cooling system (FCS) of an economized chiller. Important performance parameters include flow regulation of refrigerant flow to each 2PCP array; stability of CDU operation over zero to 100% IT load and during transient and asymmetric IT loading; and CDU stable operation at startup as server population varies from zero to full population and while hot-swapping servers from racks. Demonstration of safe operation during abnormal conditions including pump switchover and loss of heat rejection is discussed. Lessons learned are included from commissioning P2P CDU and flow conveyance including charging, flushing, and startup. Advantages of P2P in achieving sustainability goals and insights into a Total Cost of Operation (TCO) analysis of RA and refrigerant to water comparing 1P to 2P liquid cooling are discussed.