Re: 1teraflops cell processor possible?

On 1 Dec 2003 14:23:33 -0800, [EMAIL PROTECTED] (George William
Herbert) wrote:

<snip>
>
>If there is money to be made now, and science and academic
>careers to be advanced now, using today's technology and
>algorithms, why on earth should people delay working
>on the problems with today's solutions just because we
>know that eventually, hopefully, we'll be able to do it
>so much better and faster?  
>
Sometime within the last year, Andy Glew raised the question on
comp.arch of what was holding us back in the area of parallel
computation.

*No one* mentioned a lack of processing power.  The consensus was: a
lack of solid, workable *software* tools.  That on a forum dedicated
to hardware.

In a different forum, someone wrote in to ask about how he could get
his biotech matlab calculations to run faster because each was taking
a week.

I suggested he consider a cluster.  He posted back that parallel
Matlab was a subject of research, not a tool for working scientists,
that the program he was using represented man-years of development
effort, and that he had neither the skill nor the time to retool it
for use on even a two-processor machine.  He used his two processor
box to run two instances of the program, so he could get two separate
simulations done each week.

I did a little googling and found that someone had actually
implemented the program on a beowulf cluster and gave him the link.

He was amazed that someone had done such a thing and that it could be
found so quickly, but, as it turns out, the beowulf implementation I
had found was only a partially functional version of the program and
did not have the functionality he needed.

A petaflop machine would not help that poor fellow get his Ph.D.
thesis written one day sooner.

If the DOE, or anyone else, were providing a realistic level of
funding for basic research in parallel computation, I would not be so
offended at their throwing however many million at just another big
machine.

As it is, basic research on parallel computation in the US is not
being funded at anywhere near the level it should be, and I _am_
offended by projects like Blue Gene/L.

If you've got a neato idea, and it will make zillions of bucks in
biotech, go get some venture capital and do it.  If your idea is for
real, you'll find the money.

If you want to *understand* parallel computation and develop tools so
that others can do their work more easily, be prepared to make a
selfless sacrifice to humankind's fund of knowledge.

In the process of understanding how to program parallel machines, you
will, nearly for free, get alot of insight into how to build better
parallel machines.

For all the bilge and bother of this thread, I'm left with a very
basic question: can you get the energy performance out of a classical
architecture that you can get out of a streaming architecture.  Two
posters have stated without proof that they *know* that a streaming
architecture won't beat a classical architecture on realistic code.  

I'm glad that they are possessed of such a profound and instantaneous
grasp of all that is possible in computation.  I'm a little slower,
and I think most of the rest of the world is, too, and I'd like to see
a little more money go into questions like that and a lot less into
high Linpack scores.

RM