Author: Larry Woodlock
Date: 22:13:35 05/02/98
Go up one level in this thread
On a related topic, we know that the Fritz Mark varies a few
percent with each engine loading, presumably due to Windows genies. But
I've noticed something different after installing the new Fritz update
from CBM 63. On my lowly P120 with 32Mb RAM, 22 Mb hash, I have
been getting 48-52 since October, running about 88KN/s to 92KN/s. With
the update, I've gone as high as Fritz Mark 68 (!!), a 30% increase, but
the nodes have gone up between 0 and 8%.
My first look, a 68, had 97KN/s, and the record for me is now
101KN/s, with a Fritz Mark of just 57 (long). This seems to be
establishing itself as the typical high value. I'm glad for what looks
like a speed increase, but now I'm wondering. Shouldn't the Fritz Mark
value be closely tied to KN/s?
My working theory is that they tinkered a little with the Fritz
Mark values during the evolution of the program. (I have one of the
earliest). Anyone with similar experiences, or ideas on the decoupling
of Fritz Mark from node values? Thanks, Larry Woodlock.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.