[Shootout-list] Mandelbrot Set for CMUCL and SBCL 2nd try

[Yannick Gingras]

--nextPart1530580.EVEcPmDxhK
Content-Type: text/plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

On March 29, 2005 11:46 am, Greg Buchholz wrote:
> P.S. =A0A well reasoned argument might consist of stating that the
> mandelbrot set isn't well-conditioned or numerically stable. =A0But that
> doesn't address the question about the other floating point benchmarks.
> Also note that there are currently 20 programs which are able to produce
> the exact result on the mandelbrot test.
=20
As Pascal said, the other benchmarks achieved an output similar to the
C one at the cost of doing it the C way instead of doing it the way
they would normally do it.  I must add that the divergence depends on
several factors.  In example, my benchmark produce the very same
output as the C benchmark for some value of N but it gets a different
bit or two for other values.

The stability of the Mandelbrot Set increase with the number of
iterations and 50 is really small.  I suggest that we set the
iteration count to at least 1024 which is more realistic for a fractal
explorer and would make this benchmark clearly a numeric benchmark
instead of a mix of IO and numeric.  Also, if we set the number of
iterations to 1024 or more, it would be nice to decide now if we allow
cycle detections.

 ; )

=2D-=20
Yannick Gingras         Math Student, UQAM
Join FSF as an Associate Member at:
  http://member.fsf.org/join?referrer=3D2327



--nextPart1530580.EVEcPmDxhK
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQBCSoYuX5grbfhLwA4RAvS3AJ9eiV4my9nZpOuYo2lyzD/I6j4y+QCglXIZ
432yj6pDldIJmKYXBpA5AHo=
=XqO9
-----END PGP SIGNATURE-----

--nextPart1530580.EVEcPmDxhK--