YOU ARE AT:AI InfrastructureFujitsu targets lower data center energy consumption

Fujitsu targets lower data center energy consumption

Fujitsu said it will assess the effectiveness of the solution at its Tatebayashi Data Center, aiming to help customers achieve a world-class PUE at their data centers

In sum – what you need to know:

Energy efficiency push – Fujitsu is partnering with Supermicro and Nidec to develop a liquid-cooling solution that could improve data center energy efficiency by up to 40% compared to traditional air cooling.

AI-driven demand – As GPU usage for AI workloads drives up power needs, liquid cooling is emerging as a key method for reducing energy consumption.

Japan-first rollout – Fujitsu aims to launch a subscription-based offering in Japan in fiscal Q1 2025

Japanese company Fujitsu announced a collaboration with Supermicro and Nidec with the aim of improving data center energy efficiency, the former said in a release.

As part of this collaboration, the three companies will combine Fujitsu’s liquid-cooling monitoring and control software, Supermicro’s GPU servers and Nidec’s liquid-cooling system to develop a service that will help data centers increase their power usage effectiveness (PUE).

Fujitsu said it will assess the effectiveness of the solution at its Tatebayashi Data Center, aiming to help customers achieve a world-class PUE at their data centers by the fourth quarter of fiscal 2025.

“The rapid proliferation of AI is driving a surge in computing demand, leading to a global increase in the number of data centers and their overall power consumption. Higher power consumption is largely due to the increased cooling requirements of GPUs used in AI servers. Many data centers utilize air cooling, resulting in high average PUE ratios. While liquid cooling offers significantly higher efficiency, the design, construction and operation of liquid-cooling systems require a high level of expertise and skill, and deployment of these systems has been limited,” Fujitsu said.

Fujitsu explained that it will apply its expertise in liquid-cooling technology used in top-tier supercomputers and mission-critical servers around the world to develop software that provides real-time monitoring of liquid-cooled servers and optimizes cooling equipment. This is expected to reduce liquid-cooling equipment power consumption and improve overall data center energy efficiency by up to 40% compared to traditional air-cooling methods. The software will also centralize management of liquid-cooled servers and equipment, reducing data center management workload, said Fujitsu.

Supermicro will provide high-density AI server systems optimized for liquid cooling. This eliminates the need for air cooling fans, significantly reducing server power consumption and improving the server environment with reduced noise and lower operating temperatures.

Meanwhile, Nidec will provide highly efficient cooling systems centered around its Coolant Distribution Units (CDUs). Nidec will utilize its proprietary pump technology and expertise in precision engineering to deliver high-performance liquid cooling solutions. These systems will optimize thermal management within servers, reducing energy consumption and contributing to sustainable datacenter operations.

Fujitsu’s liquid-cooling monitoring and control software will also be applied in a Fujitsu offering for datacenters that will be rolled out starting in Japan in the first quarter fiscal 2025. The offering, tailored for customers who are considering liquid-cooling for their datacenters, will provide one-stop support for the entire liquid-cooling solution lifecycle, from the introduction and planning stages, to construction, operation and maintenance. Fujitsu said that this offering will be provided under a subscription model that can be deployed with low initial investment cost.

Liquid cooling is a method used to remove excess heat from electronic components by using a liquid coolant instead of air. This technique is commonly applied in high-performance computing (HPC), gaming computers and AI data centers.

There are two main types of liquid cooling:

Direct Liquid Cooling (DLC): This method involves circulating liquid directly over hot components, such as processors and GPUs, using small tubes or cold plates. The heat is transferred to the liquid, which then moves it away from the components.

-Immersion Cooling: In this approach, entire servers or components are submerged in a non-conductive cooling liquid. The liquid absorbs the heat and then carries it away through a cooling system.

ABOUT AUTHOR

Juan Pedro Tomás
Juan Pedro Tomás
Juan Pedro covers Global Carriers and Global Enterprise IoT. Prior to RCR, Juan Pedro worked for Business News Americas, covering telecoms and IT news in the Latin American markets. He also worked for Telecompaper as their Regional Editor for Latin America and Asia/Pacific. Juan Pedro has also contributed to Latin Trade magazine as the publication's correspondent in Argentina and with political risk consultancy firm Exclusive Analysis, writing reports and providing political and economic information from certain Latin American markets. He has a degree in International Relations and a master in Journalism and is married with two kids.