Growth of cloud HPC for several years has outpaced overall HPC market growth, and public cloud providers continually add the latest and greatest chips, interconnects, memory and software applications to their HPC menus. No doubt public cloud HPC is a great place for HPC experimentation and POC projects. But beyond that, how should organizations go about deciding whether to make a big move to cloud HPC – which kinds of companies and workloads should go to the cloud and which should stay on-prem?
It turns out these decisions hinge on defined characteristics of HPC workloads and the organizations running them – some of which may not be what the public cloud providers most want. This is one of the facets of cloud HPC discussed by Doug Black and Shahin Khan in this episode of the @HPCpodcast.
You can find our podcasts at our Twitter page, on insideHPC and on our blog at OrionX.net. Here’s the RSS feed: http://orionx.net/category/audio-podcast/feed
We welcome your ideas for special topics and guest commentators. Feel free to contact Doug or Shahin with suggestions.
