Gone are the days when liquid cooling for HPC-class servers (and, for that matter, other types of servers and PCs) is thought of as risky. In fact, liquid cooling has been under development for decades — it goes back much farther than you might think. And now that the power required to cool HPC clusters and data centers has become so expensive and carbon-intensive, liquid cooling has become increasingly prevalent. Some HPC centers report liquid cooling has cut their electrical spending by more than 30 percent.
These and related topics are the focus of this @HPCpodcast episode, sponsored by Lenovo. Shahin and Doug discuss liquid cooling as part of decarbonization and sustainability strategies. We cover everything from chilled doors to direct-to-chip, immersion cooling, vapor chambers, under-water data centers, how servers can be cooled by water we would normally consider to be very hot and the Brobdingnagian water pumps and pipes used to cool exascale-class supercomputers, such as the Frontier system at the Oak Ridge Leadership Computing Facility.
You can find our podcasts at insideHPC’s @HPCpodcast page, on Twitter, at the OrionX.net blog, on iTunes, and on Google. Here’s the OrionX.net podcast page, and the RSS feed.
We welcome your ideas for special topics and guest commentators. Feel free to contact Doug Black or Shahin Khan with your suggestions.