
Dell Technologies has announced the deployment of its PowerEdge XE9680L water-cooled GPU servers int...
The AMW Read
Novelty 1: confirms known infrastructure segment trajectory; Significance 2: segment-level impact for AI infrastructure as liquid cooling gains deployed middle-market reference.
Dell Technologies has announced the deployment of its PowerEdge XE9680L water-cooled GPU servers into container data centers operated by Getworks (株式会社ゲットワークス), a Japanese provider of containerized data center infrastructure powered by renewable energy. Each XE9680L can house up to eight NVIDIA GPUs and is designed for high-density liquid cooling. Getworks tested the system and reported up to 30% lower total system power consumption compared to air-cooled servers, a 1/10 reduction in air conditioning energy, and a 50% reduction in server floor space within the container. Dell's Rack Integration Service handled site survey and installation, and the iDRAC management tool integrates with Getworks' monitoring for real-time thermal oversight.
Why it matters: This deployment is a concrete proof point for the emerging 'hyperscaler distribution moat' pattern — in which large OEMs like Dell leverage global supply chains to deliver scarce liquid-cooled infrastructure faster than smaller competitors. It also updates an open debate in the AI infrastructure segment: whether liquid cooling will remain a niche for hyperscale clusters or become standard in edge and mid-market containerized data centers. Getworks plans to expand a network of virtually linked water-cooled container sites across Japan, which could signal a shift toward modular, low-power AI compute hubs outside the major cloud regions.
Grounded expert take: The article's emphasis on Dell's supply reliability as the decisive factor — "the biggest deciding factor" — confirms that liquid-cooled GPU server availability is currently constrained, giving global OEMs a structural advantage. The 50% space reduction and 40% noise drop are meaningful operational metrics for colocation and edge deployments. This is not a breakthrough in chip architecture or cooling technique, but a distribution and integration win that may accelerate adoption of liquid cooling in non-hyperscale settings. The Japanese market's land constraints and renewable-water synergy make it a natural testbed.


