[Beowulf] Cloud / HPC
ljdursi at scinet.utoronto.ca
Sun Apr 14 10:01:59 PDT 2013
I cannot agree with this piece highly enough.
Widespread cloud availability, GPU, etc, has enabled all sorts of weird, wacky, and _useful_ large-scale technical computing use cases, and arguing about whether new use case X is "really" HPC has long since lost whatever novelty it had. I'm pleased to see Jeff Layton using the broader term "Research Computing"; in my corner of the world I've been pushing for the term Advanced R&D Computing (ARC) as a catch all for any sort of technical/numerical computing that requires you to do something "special" (e.g., do something different than run naive serial code on a desktop). Someone else can probably come up with a better name, but I actually think that holding on to terms with existing pretty strong connotations is hurting more than helping at this point.
Jonathan Dursi, <ljdursi at scinet.utoronto.ca>
SciNet HPC Consortium, Compute Canada
More information about the Beowulf