Lucas Wiman writes:
Though I don't have specific timings, I imagine this would be the
case. I was referring to the Mfactor program by Peter Montgomery.
Oh. And I see he's already replied; from that info, I would guess
that Mfactor is only slightly faster than mersfacgmp on SPARCs (and
mersfacgmp can, using libgmp, factor arbitrarily far).
But actual numbers would be helpful. Volunteers? Feel free to send
numbers only to me, if you like. I already have some numbers for
MacLucasUNIX on Macs in private email that I need to put on my web
pages somewhere. Or does someone already maintain a page with timings
for non-Intel CPUs? I don't recall seeing one.
I have heard this performs considerably better than a GMP based
program (written by Alex Kruppa) on RISC architectures. I suppose
that I could/should check up on this with my SPARC-owning friend.
That would be helpful, yes.
Again, I have no timings for this, but I would think that if you tried
MacLucasUNIX on both a SPARC, and a PC, the SPARC would be the overall
winner because of the massive amount of I/O that runs on LL tests.
The only large amounts of I/O are the checkpoint files. Unless you're
doing them too often, they will be a small fraction of the run time.
Further, MacLucasUNIX, mersenne1, mersenne2, and fftlucas know about
the Intel CPU's larger floating point variables, which gives the PCs
an edge. See setup.h in the mers package for details.
In factoring, I would imagine that the difference would be smaller,
(using the same program).
This, I don't even have a real guess about; the last time I did
factoring on SPARCs, they were SPARC 2's (75 MHz?), which are much
slower than Ultra's.
Will
http://www.garlic.com/~wedgingt/mersenne.html
mers.tar.gz
mers.tgz
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
