The traditional HPC market has not embraced immersion, bypassing it in favour of direct to chip cooling technologies. The reasoning behind this is logical, as the requirements of denser compute cannot be met by immersion cooling, which is suited to more traditional applications. However, to get to the heart of the matter and fully understand the role of direct to chip cooling, we must first assess what exactly is meant by HPC.

This panel will initially dissect the terminology, breaking down any confusion by what exactly the industry means when referring to high density compute. In order to bring this analysis to life, expert speakers will assess the real life applications of direct to chip cooling and its use cases, exploring how chip architecture has shaped the face of liquid cooling so far, and what the trajectory for both high density workloads and cooling systems looks like. This panel will also broach the question, is there a limit to cold plate cooling?