[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [f-cpu] Winograd DCT on my seul.org account

----- Original Message -----
From: Yann Guidon <whygee@f-cpu.org>
To: <f-cpu@seul.org>
Sent: Monday, April 22, 2002 12:07 AM
Subject: Re: [f-cpu] Winograd DCT on my seul.org account

> > Parallel execution is non-sense for uniprocessor and even multi-processor
> > of them executes one sequential code). So I don't see the point.
> i don't have the courage to prove the contrary tonight.
> Maybe you will see the point if you do more application programming
> and see that the reason why computers are so slow is because they are so
> inherently serialised. A "parallel" langage has the purpose of showing
> what is independent from what, it doesn't force you to use a parallel
> and multithread support if an old issue in the OS community. C does not
> expose enough parallelism and requires explicit stuff which makes heavier.

I'm speaking about parallel execution per instruction, not per thread
(instructions in a thread are still serialized). So long as we get no real cpu
able to execute parallel instructions, there is no real gain for anything for a
true parallelism. Yes, a parallel language is okay to see what is independent
(the good point) and shall be surely better than C but we will still have a lot
of overhead due to the use of multithreading (stack space and switching for
example). So if a CPU has power enough, we tend to use it more in serial way
than parallel way since there is no real reason to think a multithreading would
*ALWAYS* run better.

To unsubscribe, send an e-mail to majordomo@seul.org with
unsubscribe f-cpu       in the body. http://f-cpu.seul.org/