HPC considered harmful…? [CORRECTED][UPDATED]

Print Friendly, PDF & Email

I’ve just posted a few thoughts on Greg Wilson’s “HPC considered harmful” talk, which is making its way around the interwebs these days, over at Tabor’s HPC Horizons community.

If you aren’t signed up as a Horizons community member, I’d encourage you to take a moment and register. It’s free, and registration will give you the opportunity to sound off on discussions posted there, and to post your very own discussion on a topic, question, or issue of interest to you.

Speaking of integrity, now I get to practice what I preach. Sorry for the totally bogus info on this, and thanks to Mike Heinz for pointing out that I needed to do a little more research. Horizons is an interesting community, and I would absolutely encourage you to get in touch with them to learn more, but it isn’t free. There are corporate and individual memberships available. Talk to Tabor to find out more.

[UPDATE: JUST IN – As a special offer to insideHPC readers, Tabor Communications will offer a complimentary one-year membership to HPC Horizons (valued at $395). If you’re interested, you’ll first need to sign in and register online at www.hpcwire.com. Type “insideHPC” into the Promotion Code box and be sure to request HPC Horizons. If you have any questions, you can contact debbie@taborcommunications.com directly.]


  1. […] HPC considered harmful…? [CORRECTED] […]


  1. One of my professors at MSU, Dr. Timothy Swafford, always said that the worst answer is the answer you get that looks correct.

  2. Mike Heinz says

    Signing up might be free, but trying to read your article seems to be a pay-only proposition.

    Kind of a bait-and-switch, don’t you think?

  3. Mike – as we’ve talked about by email, I’m working with the Horizons guys to figure out if joining really isn’t free, or if the site is having issues.

  4. If the article does become truly free, I’d be interested in reading it as well (for obvious reasons 🙂

  5. I thoroughly enjoyed reading Greg’s paper, and will be sending the link to a number of people. I had to fight tooth and nail to get my place of work to purchase a debugger a while back, and now we’re fighting over compiler licenses, too. There’s a sense amongst people new to the field that high performance computing is merely throwing hardware at a problem, and hoping things work. The notion of setting aside a mere 2-3% of the hardware funds for software is considered heresy, despite the enormous returns on investment in the hardware one receives from doing so!

    …. And I haven’t even started ranting yet on the code development process, validating results, training, optimization, etc. 🙂

    Anyways, thanks to Greg for writing this, and to John for posting it, and to HPCwire / Tabor for the free registration for the year!

  6. Why would you need a debugger for the Top500 benchmark? 😉