Liquid cooling solutions gain momentum as artificial intelligence workloads raise temperatures in data centers

When you make a purchase through links in our articles, Future and its syndication partners may earn a commission.

Artificial intelligence is driving unprecedented demand for data centers as the need to process massive amounts of data continues to grow.

As tech giants look to expand their infrastructure to support AI workloads, they face a growing challenge of how to power those operations sustainably and affordably—and it’s even prompted companies like Oracle and Microsoft to explore nuclear power as a potential solution. .

Another important issue is managing the heat generated by powerful AI hardware. Liquid cooling has emerged as a promising way to maintain optimal system performance while meeting increasing energy demands. In October 2024 alone, several technology companies announced liquid cooling solutions, highlighting a clear industry shift in this direction.

Liquid-cooled superclusters

At the recent Lenovo Tech World event, the company showcased its next-generation Neptune liquid cooling solution for servers.

The sixth generation of Neptune, which uses direct open-loop warm water cooling, is currently being deployed across the company’s partner ecosystem, enabling organizations to build and run accelerated compute for generative AI while reducing data center energy consumption by up to 40%. , – the company reports.

At OCP Global Summit 2024, Giga Computing, a subsidiary of Gigabyte, introduced a direct liquid cooling (DLC) server designed for Nvidia HGX H200 systems. In addition to the DLC server, Giga also introduced the G593-SD1 with a dedicated air cooling chamber for the Nvidia H200 Tensor Core GPU, aimed at those data centers that are not yet ready to fully embrace liquid cooling.

The new Dell Integrated Rack 7000 (IR7000) is a scalable system designed specifically for liquid cooling. It is capable of managing future installations up to 480 kW while capturing almost 100% of the heat generated.

“Today’s data centers cannot meet the demands of artificial intelligence, requiring high-density compute and liquid cooling innovations with modular, flexible and efficient designs,” said Arthur Lewis, president of Dell’s Infrastructure Solutions Group. “These new systems provide the performance organizations need to remain competitive in the rapidly evolving artificial intelligence environment.”

Supermicro also introduced liquid-cooled SuperClusters designed for AI workloads and running on the Nvidia Blackwell platform. Supermicro’s liquid cooling solutions, supported by Nvidia’s GB200 NVL72 exascale computing platform, have begun sampling to select customers, with full production expected late in the fourth quarter.

“We are creating the future of sustainable computing powered by AI, and our liquid-cooled AI solutions are rapidly being adopted by some of the most ambitious AI infrastructure projects in the world, with over 2,000 liquid-cooled racks delivered since June 2024,” said Charles Liang, President and CEO of Supermicro.

Liquid-cooled superclusters feature advanced rack- or row-mounted coolant distribution units (CDUs) and custom cooling fins to accommodate two Nvidia GB200 Grace Blackwell superchips in a 1U form factor.

It seems clear that liquid cooling will become the mainstay of data center operations as workloads continue to grow. This technology will be critical to managing the heat and energy needs of the next generation of AI computing, and I think we are just beginning to see the potential impact it will have on efficiency, scalability, and sustainability in the coming years.

More from TechRadar Pro