Why Obama Is Wrong with His Super Computer Decision

2 points by Ryanb58 9 years ago | 8 comments
  • natch 9 years ago
    Obama just signed the order. NASA people and presumably other technical people were involved in the decision.

    It does seem like something of an interesting and worthwhile challenge to conduct a project like this while performance is being lifted ever higher in step with Moore's law, or, if it is not sustained, whatever curve follows it.

    • Ryanb58 9 years ago
      More than likely. I am just trying to get them to think outside of the specialized box they are stuck in.

      With the wide spread of fiber cable(faster internet speeds), I see it being very do-able form of computing in the near to soon future. We might as well help push the Moore's law mentality a bit further.

    • GFK_of_xmaspast 9 years ago
      Ryan, do you know anything about the high performance computing environment.
      • Ryanb58 9 years ago
        Indeed I do. It'd of been kind of silly of me to post something about it if not. LOL

        The goal is to get people thinking outside of the box. Especially those in leadership positions.

        • GFK_of_xmaspast 9 years ago
          I ask because, in my experience, one of the things that comes along with HPC is 'a ton of data that has to move all over the data at high speed'.

          Your proposal is to formalize the folding@home and similar efforts into a national distributed general-purpose computing environment. This kind of system works fine if the problem can be decomposed into small independent computation-bound pieces, but that's not a framework that fits a lot of things.

          You write: "To get a program working on a super computer, one must pay highly specialized developers in order to create a program that might only be used once. It also might take a long time to build considering the missing tools which are available to use when targeting the consumer side of the market."

          I don't see any reason why this would not hold in any other computing environment, and tbh I'd expect it would be even worse. We already have computational power on desktops that would have been unimaginable not that long ago, not to mention GPUs, but look how shoddy the scientific computing ecosystem is (there are exceptions, but in general it's "not good").

      • markhahn 9 years ago
        no content in your complaint, just your various dislikes.
        • dynomight 9 years ago
          Why not have both?
          • Ryanb58 9 years ago
            Think of it like using a coupon... Why spending $100 dollars on a pair of shoes, when you can use a coupon and save 50%.

            Plus America is in a huge amount of debit.

            Not saying we couldn't do both ideas though. I wonder which would be more useful.