On Mar 25, 1:33 pm, Bruce McFarling <agil...@[EMAIL PROTECTED]
> wrote:
> On Mar 25, 1:09 pm, an...@[EMAIL PROTECTED]
(Anton Ertl)
> wrote:
>
> > I am not an expert in Grid Computing, but I have heard a number of
> > talks about it, and my impression is that the problems that they are
> > working on are very different from what you would ever encounter with
> > the kind of machine you are thinking of; and vice versa, in a SeaForth
> > chip you encounter issues that you will not encounter in a
> > supercomputing node.
> > ...
> > I also believe that it is easier to develop, say, some wheather
> > prediction code by starting on a PC than by starting on a 5x8 grid of
> > SeaForth chips
>
> Yes, but the issue raised by Jeff M.'s post was whether there was
> reason to be concerned that there was no Forth programming model that
> would be workable for massively superparallel computing.
Yes. And in small-scale parallel processing (but across multiple cores
- not just threads).
> And given that the source of concern is a purely hypothetical one, I
> see nothing wrong with addressing it with a purely hypothetical system
> composed of an array of modules consisting of 5x8 arrays of SeaForth
> chips.
True.
But, while the question was phrased hypothetically, it's obviously not
a hypothetical problem. We are quickly moving into an era where faster
computers aren't the solution, but rather multiple computers. My
particular area of expertise is in console games. In the past (Atari -
> Nintendo -> Genesis -> Xbox), it was all about getting faster
processors. With the recent generation of consoles, 90% of the work I
do is in multi-core scheduling, processing, sharing, ...
The Xbox 360 has 6 cores and the PS3 has a 2 core PPU and multiple SPU
cell processors. We've had to completely rethink game programming. We
have physics running on GPUs, rendering on SPUs, data reading off disk
on the PPU, etc, etc. And this trend is only going to continue.
Forth is a beautiful language, and while it _can_ be used to solve any
problem out there, just like Pascal could, some languages are just not
suited to solve all problems with equal ease. I seriously cannot
picture myself solving the kind of problems I'm solving daily (in
regards to job scheduling, timing memory accesses, ...) in Forth just
as I can't picture myself writing embedded code in Pascal. But,
perhaps that's because I've just never tried.
Please keep in mind that this isn't a bash on Forth. I love the
language. I just wondered if there were attempts at making Forth work
on more "modern" hardware: vector processors, IBM's new CELL,
accessing non-unified memory, and similar processors where parallel
computing is the win - not just faster processors.
I enjoyed Elizabeth's reply and reading about SeaForth.
Jeff M.
|
