Author: Slater Wold
Date: 21:46:39 10/11/01
Go up one level in this thread
On October 12, 2001 at 00:35:44, K. Burcham wrote: > >these are two examples of slates upgrade from 2x1400mhz, to 2x1533mhz. > > >out of the three tests, one of the tests had the analysis shown at different > times and different depths, didnt seem to be a good comparison. That's the beauty of SMP. :) > >but the two tests below are close enough to see a difference, since the >time and depth were pretty close in the analysis. I already did this, for myself. But I guess this shows everyone here the difference. > >i added the kns. value. first test gain of 137kns. > second test gain of 171kns. first test was accessing egtbs Crafty gained 100k nps. And the EGTB's slow down nps. (Obviously.) It's roughly a 9% gain. (According to the tests I've ran.) > > > >2800 mhz > >+- (3.02) Depth: 27 00:02:02 169000kN, tb=6750 >1.b7 Re8 2.Ba7 Be1 3.b8Q Rxb8 4.Bxb8 Bb4 5.Bd5 Kh5 6.Bf7+ Kh6 7.Bc7 > +- (3.02) Depth: 29 00:07:10 672750kN 1564kns tb=28655 > >3066 mhz > > +- (3.02) Depth: 27 00:01:53 169393kN, tb=6762 >1.b7 Re8 2.Ba7 Be1 3.b8Q Rxb8 4.Bxb8 Bb4 5.Bd5 Kh5 6.Bf7+ Kh6 7.Bc7 Kg7 8.Bd5 >Kf6 > +- (3.02) Depth: 29 00:06:36 673858kN 1701kns tb=28699 > > > > >2800 mhz > >1.Nxg5 Bxd1 2.Nxe6 Qb8 3.Nxg7+ Kd8 4.Kxd1 b5 5.Bf5 b4 6.Ne6+ Kc8 7.Bd2 Qb6 8.Kc2 > ² (0.68) Depth: 18 00:03:52 363925kN >1.Nxg5 Bxd1 2.Nxe6 Qb8 3.Nxg7+ Kd8 4.Kxd1 b5 5.Bf5 b4 6.Kc2 a5 7.Ne6+ Kd7 8.Bh6 >a4 9.cxb4 > ± (0.72) Depth: 20 00:07:17 757031kN 1732kns > >3066 mhz > >1.Nxg5 Bxd1 2.Nxe6 Qb8 3.Nxg7+ Kd8 4.Kxd1 b5 5.Bf5 b4 6.Ne6+ Kc8 7.Bd2 Qb6 8.Kc2 > ² (0.68) Depth: 18 00:03:32 361911kN >1.Nxg5 Bxd1 2.Nxe6 Qb8 3.Nxg7+ Kd8 4.Kxd1 b5 5.Bf5 b4 6.Kc2 a5 7.Ne6+ Kd7 8.Bh6 >a4 9.cxb4 > ± (0.72) Depth: 20 00:06:54 787819kN 1903kns > >(thanks to slate) No problem. >kburcham
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.