Actually, I just wanted to use the word “epoch” in a post.
Anyway, good stuff is going on over at HPCwire. Tabor Communications has reorganized under the leadership of Debra Goldfarb, a name many of you will recognize. The new company-wide focus will be on helping the community build a conversation around the value of supercomputing. As Michael Feldman put it in last week’s article:
That’s going to make my job a bit more complex. Unlike simple computing metrics like SPEC or Linpack benchmarks; network measurements of bandwidth and latency; or storage measurements of capacity and data transfer rates; productivity is notoriously hard to quantify. It’s one of those things that people think “I know it when I see it,” but it’s difficult to measure.
Of course these issues aren’t new, we are just newly focusing on them. I’m excited by the change. [Full disclosure: I write a weekly (non-compensated) career feature for HPCwire.]
A better focus on overall productivity and value delivered is a great sign of a maturing field – I’d say academic HPC research is lagging behind in this respect, because it is so much more comfortable to focus on hard performance numbers than to address problems that make high performance so hard to achieve, or make computation a small part of the overall HPC task.
If anyone is interested in academic attempts to quantify overall productivity, take a look at the work of the DARPA HPCS Productivity Team, found in this issue of CTWatch: http://www.ctwatch.org/quarterly/archives/november-2006-a/ , and in this issue of IJHPCA:
http://www.highproductivity.org/IJHPCA.html
(Full Disclosure, I’m a part of the HPCS team, and my own research involves HPC User Productivity.)