HPE launches fanless systems architecture for AI

Hewlett-Packard Enterprise has launched its fully fanless direct liquid-cooled systems architecture to support large-scale AI deployments. The company says its new innovation aims to significantly reduce power consumption to support organisations needing to run large AI workloads.

The architecture uses the same liquid cooling technology HPE uses for seven of the top 10 supercomputers on the Green500 list of the world’s most energy-efficient supercomputers.

The system architecture is built on an eight-element cooling design that includes liquid cooling for the server blade, network fabric, GPU, CPU, rack or cabinet, pod or cluster and coolant distribution unit.

The design offers benefits including a 37% reduction in cooling power required per server blade when compared to hybrid direct liquid cooling, according to HPE. Because systems utilising the architecture can support greater server cabinet density, they need only half the floor space.

HPE President and CEO Antonio Neri said today’s organisations are under pressure to simultaneously embrace the power of generative AI while advancing sustainability goals and lowering operational costs.

“The architecture we [have] unveiled uses only liquid cooling, yielding greater energy- and cost-efficiency advantages than the alternative hybrid cooling solutions on the market,” he said. “HPE’s expertise in deploying the world’s largest liquid-cooled IT environments and our market leadership spanning several decades put us in excellent position to continue to capture AI demand.”

Image credit: iStock.com/denisik11