Re: tlug: Karl-Max has cool dreams [was: dual-pentium processors]

["Manuel M. T. Chakravarty" <chak@example.com>]

"Stephen J. Turnbull" <turnbull@example.com> wrote,

> >>>>> Manuel M T Chakravarty <chak@example.com> writes:
> 
>     >> Who needs to?  `M-! ps ax | wc' ==> `54 320 2233'
>     >> 
>     >> I've got 54 processes running,
> 
>     Manu> Yeah, 95% of them sleeping.  Won't help you much utilizing
>     Manu> your CPUs.  A better measure is the CPU load of your machine
>     Manu> (i.e., average number of processes in the ready-to-run
>     Manu> queue).  I guess you are not really often above 1, 2, or
>     Manu> maybe 3 processors.
> 
> Never, unless I'm running a compilation and cthugha at the same time,
> but note the cost assumption I made: $/processor then = $/byte today.
> Also, I'm running what is basically a glorified word processor.  I
> don't use multimedia (especially not audio/video input), do image
> processing, etc, etc.  But suppose it becomes cheap to put a processor 
> on each of those tasks?

I am sorry, but actually, your approach of usin many small
and cheap processor was already tried out.  The company was
Thinking Machine and the machine the CM-2.  The company had
to face Chapter 5 a couple of years ago.

Especially, as writing scalable programs is difficult, it
turned out that it is more efficient to use few high-end
processor than many low- or mid-range processor.

> Also, there are lots of processor intensive tasks that don't get done
> now because we don't have the technology, but will soon.  Voice
> recognition, for example.  Will we find ways to use 64 microprocessors
> at a time?  Well, how many microprocessors does the typical late-model
> car have?

A car is a distributed processing system, not a parallel
system.

> 
>     Manu> For some applications, you can use many processors no
>     Manu> question, but without a revolution in programming languages,
>     Manu> compilers, and such it will be a few carefully selected and
>     Manu> handcrafted pieces of software. (And the so-called
>     Manu> `embracingly parallel problems' like computing a Mandelbrot
>     Manu> set).  But maybe there will be a revolution :-)
> 
> How about the game of life?

This belongs to the embracingly parallel problems.

> Or page-at-a-glance OCR to give a practically important
> example.

Actually not that easy to parallize, because when it comes
to the high-level stages like feature recognition etc.

Cheers,

Manuel
--------------------------------------------------------------
Next Nomikai: 18 September, 19:30 Tengu TokyoEkiMae 03-3275-3691
Next Meeting: 10 October, Tokyo Station Yaesu central gate 12:30
--------------------------------------------------------------
Sponsor: PHT, makers of TurboLinux http://www.pht.co.jp