Some realities that seem to be missing from this discussion.

One is that the big bucks for parallel computing in the US come from "scientific" computing (translation: "scientific" = "bomb"). Make some friends at LANL or LLNL. Surely you and the parallel computing "guru" from NASA must have met.

The second reality is: much money, little insight. I've already made enough enemies, but the dwarfs in this area are not the problems. The dwarfs in this area are comfortably-situated power brokers who made their careers on "solutions" that were both obvious and limited (lots of cheap hardware).

No more whompti-flops computers needed. No more ad hoc languages. What's needed is the kind of work the US doesn't currently fund: basic research without an immediate obvious payoff. Mathematicians welcome. Bureaucrats and computer "scientists" please go away.

Robert.


Linus Torvalds (torvalds@osdl.org) on 2/14/08 wrote:
---------------------------
>Ugh. They seem to make essentially all of their arguments
>based on their "dwarfs" (shouldn't that be "vertically
>challenged algorithm"?).
>
>And their dwarfs in turn seem entirely selected to then
>support the end result they wanted. Can anybody say
>"circular argument" ten times fast?
>
>Apart from the obvious graphics thing, none of their loads
>seem at all relevant to "general purpose computing", they
>are all essentially about scientific computing.
>
>And we already pretty much know the solution to scientific
>computing: throw lots of cheap hardware on it (where "cheap"
>is then defined by what is mass-produced for other reasons).
>
>Designing future hardware around the needs of scientific
>computing seems ass-backwards. It's putting the cart in
>front of the horse.
>
>Linus