Loading ...
Sorry, an error occurred while loading the content.

RE: [decentralization] Distributed Computing Economics

Expand Messages
  • Werner Vogels
    ... I think the main strength of the article is that he shows for what class of applications WAN style Grid computing can possibly succeed: computations must
    Message 1 of 2 , Jul 8, 2003
    • 0 Attachment
      > Database pioneer Jim Gray explains why all
      > the "distributed computing" startups from
      > 2001 failed: it simply isn't economical.

      I think the main strength of the article is that he shows for what class
      of applications WAN style Grid computing can possibly succeed:
      "computations must be nearly stateless and have more than 10 hours of
      cpu time per GB of network traffic before outsourcing the computation
      makes economic sense."

      > One thing he doesn't address in his mention of
      > SETI@Home is the psychological aspect of incentives:
      > if you give someone a cool screensaver in exchange for
      > their spare CPU cycles they'll think it's cool; if you
      > give them a dollar they'll most likely consider it
      > an insult.

      I think an even stronger social incentive in the SETI@Home case was the
      competition issue, people would suddenly start investing in more and
      better hardware, just to get their ratings up at the SETI site.

      --
      Werner
      http://weblogs.cs.cornell.edu/AllThingsDistributed
    Your message has been successfully submitted and would be delivered to recipients shortly.