Comparison of multiple-precision floating-point software

paul zimmermann Paul.Zimmermann at inria.fr
Wed Apr 25 09:58:26 UTC 2018


       Dear Torbjörn,

> I am surprised that there are non-marginal differences for larger
> operations.  Don't we all use mpn?  Bookkeeping should be the only
> difference.

since 2004 we did implement in MPFR "short" product/square/division,
which compute an approximation of the upper n limbs of a nxn product
or square, or of the quotient of a 2n/n division, following this paper:

@article{Mulders00,
  author =       {T. Mulders},
  title =        {On short Multiplications and Divisions},
  journal = "AAECC",
  year = 2000,
  volume = 11,
  number = 1,
  pages = "69--88"
}

If such algorithms would be available in GMP, of course we would use them.

Paul


More information about the gmp-devel mailing list