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

Re: [f-cpu] RC5, F-CPU and srotl

On Sun, Apr 14, 2002 at 01:00:58AM +0200, cedric wrote:
> What I want to do by the port of the dnetc client is to see where we can have 
> problem with our instruction set on high parellisable algorithm. The 
> conclusion of this discussion is that we will have a "real" simd rotl, but 
> not because of this algorithm, but because we want a "clean" CPU.

Who is "we" in this case?

I tend to agree with you, but in general, we have to consider carefully
what we include and how we do it. If the cost/benefit ratio of a
particular feature is too high, we'd better drop it.

> > What's a 64-bit CPU good for, compared to a 32-bit one? It's not
> > necessarily faster, but it can handle large working sets. Therefore
> > I'd rather focus on scientific applications, databases and so on.
> Yes, but we need to start from a point, if you have some idee of which 
> algorithm or part of program to port, say it here, and if some one have time 
> he can start to port them.

In assembler?

> > > The case of dnet is looking different but i can't involve myself
> > > in this now. The case of the SIMD shift has been proved and there
> > > does not seem to be any difficulty in implementing it with F-CPU,
> > > we simply disagree on a "small" implementation detail :-)
> > Huh? No difficulty?
> I see your last post, so you find a solution, no ? ;-)

Yes. But I'm still unsure what it will cost (in terms of delay time and
die area).

 Michael "Tired" Riepe <Michael.Riepe@stud.uni-hannover.de>
 "All I wanna do is have a little fun before I die"
To unsubscribe, send an e-mail to majordomo@seul.org with
unsubscribe f-cpu       in the body. http://f-cpu.seul.org/