Cooling down the heat: Why liquid cooling is now mission-critical for AI datacenters

As enterprise demand for AI and high-performance computing accelerates, the infrastructure supporting these workloads is generating heat at levels that conventional air cooling simply cannot manage.

A new IDC InfoBrief, sponsored by Lenovo and based on a global survey of 1,230 IT decision-makers, finds that spending on AI and HPC workloads is expected to grow at an average rate of 52% over the next 12 to 24 months — nearly double the rate of general computing infrastructure. That growth is placing acute pressure on datacenter cooling strategies.

IDC projects global datacenter energy consumption will nearly triple by 2028, reaching 915 terawatt hours. With cost, complexity, and integration barriers still slowing adoption, the research makes clear that IT leaders who invest now, with the right OEM partner guiding strategy, design, and deployment; will be best positioned to scale AI infrastructure reliably and sustainably.

Read below to find out more.