Is there a platform optimized version of gmp for INTEL?
Michael.Abshoff at fsmath.mathematik.uni-dortmund.de
Mon Jan 1 14:35:57 CET 2007
Patrick Pelissier wrote:
> It is MPFS, not MPFR. MPFR team organised this session so they
> couldn't participate.
> You should look at the previous session for a comparison with MPFR.
you are correct. It was late and I didn't look too closely. I initially
looked at the 2005 results because it was the first hit on google when
searching for "Many digits competition", but I thought why not provide the
results from the last meeting in 2006.
According to the MPFS homepage the few tests where they publish benchmarks
those two seem to be roughly equal in speed (as long as you choose the
"right" amount of precision. Overall the MPFS looks like a very
(Because we are getting off topic I will shut up now :)
> Patrick Pelissier
More information about the gmp-discuss