SoftBank Corp., ZutaCore and Hon Hai Expertise Group (“Foxconn”) have applied ZutaCore’s two-phase DLC (Direct Liquid Cooling) expertise in an AI server utilizing NVIDIA accelerated computing, making it the world’s first implementation of ZutaCore’s two-phase DLC (Direct Liquid Cooling) expertise utilizing NVIDIA H200 GPUs. As well as, SoftBank designed and developed a rack-integrated resolution that integrates every part of the server, together with cooling tools with two-phase DLC expertise, on a rack scale, and carried out an operational demonstration and efficiency analysis at its knowledge middle in February 2025. The demonstration outcomes indicated the answer handed NVIDIA’s temperature check (NVQual), thereby confirming the compatibility, stability and reliability of this rack-integrated resolution. The answer additionally achieved pPUE (partial Energy Utilization Effectiveness) of 1.03 (precise measured worth) per rack for cooling effectivity.
With the unfold and elevated adoption of AI, demand for AI servers and different computing sources is predicted to develop considerably, additional rising energy consumption at knowledge facilities. On the similar time, lowering energy consumption from the attitude of lowering carbon dioxide (CO2) emissions has turn into a urgent concern worldwide, requiring knowledge facilities to turn into extra environment friendly, eat much less energy, and introduce progressive warmth elimination options. Since Might 2024, SoftBank has been collaborating with ZutaCore, a world chief within the improvement and enterprise deployment of two-phase DLC expertise, to develop options optimized for low energy consumption at AI knowledge facilities.
SoftBank, ZutaCore and Foxconn will proceed their efforts to attain larger effectivity and decrease energy consumption in AI knowledge facilities utilizing NVIDIA accelerated computing, and collaborate on the event of rack-integrated options optimized with two-phase DLC expertise for sensible use and deployment within the world market.
Options of rack-integrated resolution optimized with two-phase DLC expertise
1. Plug & Play resolution
By integrating server elements similar to cooling tools outfitted with 2-phase DLC expertise, which conventionally needed to be optimized individually, on a rack scale, it’s attainable to supply a “Plug & Play”*4 resolution that can be utilized instantly by merely connecting the tools. This makes it simple and easy to put in in AI knowledge facilities.
2. Revolutionary 2-phase DLC cooling system
ZutaCore’s two-phase DLC expertise, which cools the chipsets (CPUs, GPUs, reminiscence, and so on.) which might be the warmth sources within the server by bringing them into contact with a chilly plate, circulates a particular insulating warmth switch (dielectric) fluid with a low boiling level within the chilly plate and makes use of warmth absorption in the course of the section change from liquid to gasoline attributable to the warmth supply. Environment friendly cooling is achieved by means of repeated evaporation and condensation. As well as, the water temperature will be maintained at the next temperature than in standard water-cooled amenities, which improves the warmth transport capability of the water used for cooling and reduces the circulate fee of water to attain the required cooling impact. This makes it attainable to scale back the strain of the pump that delivers the water for cooling, contributing to decrease energy consumption. Moreover, the usage of insulating warmth switch fluid prevents critical injury to the server within the occasion of a spill of particular fluid.
3. AI server with assured compatibility, stability and reliability
Foxconn developed an AI server based mostly on NVIDIA H200 GPUs, optimized with ZutaCore’s two-phase DLC expertise. This AI server was capable of exceed goal efficiency figures in NVIDIA’s certification check (NVQual), which determines the compatibility, stability, and reliability of server programs.
4. Versatile and scalable rack for high-density placement of AI servers
SoftBank developed a rack designed to maximise cooling effectivity and enhance operational effectivity with its two-phase DLC expertise. Along with being designed to be versatile and scalable for high-density placement of the most recent AI servers, the rack has airflow for improved cooling effectivity, and energy provide. As well as, energy provide and wiring are concentrated within the rear of the rack for operational security and robot-friendly operation. The rack can be appropriate with 21-inch servers and standard 19-inch servers compliant with the ORV3 commonplace, which is a specification set by the Open Compute Challenge (OCP), a non-profit group that promotes open-source {hardware} specs and designs. The specs are designed to be deployed in accordance with world requirements.
Future Developments
SoftBank, ZutaCore, and Foxconn will strengthen cooperation with AI server and rack design firms to enhance the effectivity and low energy consumption of AI knowledge facilities utilizing NVIDIA accelerated computing, with a view to the commercialization of this resolution and its deployment within the world market.
