Author: Roy Eassa
Date: 16:29:57 01/26/02
Go up one level in this thread
On January 26, 2002 at 19:24:11, Roy Eassa wrote: >On January 25, 2002 at 15:51:34, Ernst Walet wrote: > >>On January 25, 2002 at 02:37:48, Ernst Walet wrote: >> >>>On January 24, 2002 at 16:27:22, Roy Eassa wrote: >>> >>>>On January 24, 2002 at 15:58:27, Ernst Walet wrote: >>>> >>>>>On January 24, 2002 at 15:14:00, Roy Eassa wrote: >>>>> >>>>>>On January 23, 2002 at 22:04:13, Dan Andersson wrote: >>>>>> >>>>>>>[D]r2qkb1r/pb3p1p/1pp1p1pB/8/3PB1Q1/6P1/PPP2P1P/R3R1K1 w kq - 0 15 >>>>>>>How does you favourite program do? Does it find 15. Bxc6+ in reasonable time? >>>>>> >>>>>>Fritz 7.0.0.6 on Athlon 1.2 takes over 6 minutes(!): >>>>>> >>>>>>1.Bg5 Be7 2.Bxe7 Qxe7 3.Qf3 0-0 4.Rad1 Qb4 5.b3 Qa3 6.Ra1 Qb2 7.c3 Rac8 >>>>>> ± (0.79) Depth: 10/26 00:00:02 1270kN >>>>>> ² (0.63) Depth: 15/36 00:03:35 136212kN >>>>>>1.Bxc6+ Bxc6 2.Rxe6+ fxe6 3.Qxe6+ Be7 4.Qxc6+ Kf7 5.Qc4+ Ke8 6.Re1 Rc8 7.Rxe7+ >>>>>>Kxe7 8.Bg5+ >>>>>> ² (0.66) Depth: 15/38 00:06:06 229372kN >>>>>> +- (1.85) Depth: 16/38 00:07:11 271799kN >>>>> >>>>> >>>>>Roy, what hashtable size did you use? With 512MB Fritz needs 400214kN, >>>>> >>>>>1.Bxc6+! >>>>> ² (0.69) Depth: 15/41 00:10:07 400217kN >>>>> >>>>>and with 832MB hash Fritz needs 353330kN. >>>>> >>>>>1.Bxc6+! >>>>> ² (0.69) Depth: 15/39 00:08:56 353330kN >>>>> >>>>>Ernst. >>>> >>>> >>>>Ernst, I use 384 MB hash. I tested Fritz 7.0.0.6. A bit of the time it was >>>>running in the background (behind the IE6 browser), so the times probably should >>>>have been a bit shorter. I'm running WinXP on an Athlon 1.2 GHz. What are the >>>>specs of your setup? >>> >>> >>>I have a Celeron 1200 with 1280MB ram, I usually use 832MB for hash and was >>>suprised that I needed more nodes than you, so I suspected that you used a very >>>large hashtable. When I get home, I will try it with 384MB hash and report >>>again then. I also run my system on WinXP. >>> >>>Ernst. >> >>Well, I ran the test with 384MB hash but need more nodes than you do, although >>it was less than with 512 and 832 MB hash. Don't now why though. >> >>Ernst. > > >OK, I tried again. This time it took Fritz 7.0.0.6 (XP, 1.2 GHz) almost 13 >minutes. > >Analysis by Fritz 7: > >1.Bg5 Be7 2.Bxe7 Qxe7 3.Qf3 0-0 4.Rad1 Qb4 5.b3 Qa3 6.Ra1 Qb2 7.c3 Rac8 > ² (0.63) Depth: 15/34 00:04:03 138340kN >1.Bf4! > ² (0.66) Depth: 15/39 00:09:33 317087kN >1.Bxc6+! > ² (0.69) Depth: 15/40 00:12:49 425842kN >1.Bxc6+! > ± (0.85) Depth: 15/40 00:12:52 427661kN >1.Bxc6+! > ± (1.16) Depth: 15/40 00:13:16 441386kN >1.Bxc6+! > +- (1.79) Depth: 15/41 00:13:43 455996kN >1.Bxc6+ Bxc6 2.Rxe6+ fxe6 3.Qxe6+ Be7 4.Qxc6+ Kf7 5.Qc4+ Ke8 6.Re1 Rc8 7.Rxe7+ >Kxe7 8.Bg5+ > +- (1.85) Depth: 16/40 00:14:10 471919kN I was focusing on time. Looking at nodes versus cache size, we see: Cache Nodes ----- ----- 384M 425842k 512M 400214k 832M 353330k I guess this makes sense.
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.