Author: Slater Wold
Date: 05:39:46 09/16/01
Go up one level in this thread
On September 16, 2001 at 07:20:40, Uri Blass wrote: >On September 16, 2001 at 06:20:38, Slater Wold wrote: > >>On September 16, 2001 at 05:40:31, Uri Blass wrote: >> >>>On September 15, 2001 at 18:37:06, Slater Wold wrote: >>> >>>>On September 15, 2001 at 17:12:37, Vincent Diepeveen wrote: >>>> >>>>[D]8/p4bpk/7p/3rq3/3Npp2/PPQ3P1/3R1PKP/8 w - - 3 32 >>>> >>>>Deep Fritz fails low on gxf4 in :54 seconds, but has a hard time replacing it. >>>> >>>>Deep Fritz T29c - W,S >>>>8/p4bpk/7p/3rq3/3Npp2/PPQ3P1/3R1PKP/8 w - - 0 1 >>>> >>>>Analysis by Deep Fritz T29c: >>>> >>>>1.gxf4 >>>> ³ (-0.50) Depth: 1/3 00:00:00 >>>>1.Kg1 fxg3 >>>> ³ (-0.44) Depth: 2/14 00:00:00 >>>>1.Rd1 f3+ >>>> ³ (-0.28) Depth: 2/14 00:00:00 >>>>1.Rd1 f3+ >>>> ³ (-0.28) Depth: 2/14 00:00:00 >>>>1.Rd1 f3+ >>>> ³ (-0.28) Depth: 2/14 00:00:00 >>>>1.Rd1 f3+ >>>> ³ (-0.28) Depth: 2/14 00:00:00 >>>>1.Rd1 f3+ >>>> ³ (-0.28) Depth: 2/14 00:00:00 >>>>1.Rd1 f3+ >>>> ³ (-0.28) Depth: 2/14 00:00:00 >>>>1.Rd1 f3+ >>>> ³ (-0.28) Depth: 2/14 00:00:00 >>>>1.Rd1 Qh5 2.Qe1 f3+ 3.Kg1 >>>> ³ (-0.41) Depth: 3/13 00:00:00 3kN >>>>1.h3! >>>> ³ (-0.37) Depth: 3/15 00:00:00 4kN >>>>1.h3! f3+ >>>> ³ (-0.34) Depth: 3/15 00:00:00 5kN >>>>1.gxf4! >>>> ³ (-0.31) Depth: 3/16 00:00:00 6kN >>>>1.gxf4 Qxf4 2.Qe3 Qg4+ 3.Qg3 Qh5 >>>> = (-0.25) Depth: 4/14 00:00:00 7kN >>>>1.gxf4 Qxf4 2.Qe3 Qg4+ 3.Qg3 Qh5 4.Kg1 >>>> ³ (-0.31) Depth: 5/15 00:00:00 20kN >>>>1.gxf4 Qxf4 2.Qe3 Qg4+ 3.Qg3 Qh5 4.Kg1 a6 >>>> ³ (-0.31) Depth: 6/17 00:00:00 41kN >>>>1.gxf4 Qxf4 2.Qe3 Qg4+ 3.Qg3 Qh5 4.Qe3 Qg6+ 5.Qg3 Qf6 >>>> ³ (-0.31) Depth: 7/19 00:00:00 77kN >>>>1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kh1 Qg4 4.f4 exf3 5.Nxf3 Bd5 >>>> ³ (-0.37) Depth: 8/22 00:00:00 163kN >>>>1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kh1 Qg4 4.f4 exf3 5.Nxf3 Bd5 6.Rf2 >>>> ³ (-0.34) Depth: 9/21 00:00:00 435kN >>>>1.gxf4-- >>>> ³ (-0.66) Depth: 10/28 00:00:00 829kN >>>>1.gxf4-- Qxf4 2.Qe3 Rg5+ 3.Kf1 Qxh2 4.Qxe4+ Bg6 5.Qe3 Re5 6.Qf3 Be4 >>>> µ (-0.91) Depth: 10/28 00:00:01 1322kN >>>>1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kf1 Qxh2 4.Qxe4+ Bg6 5.Qe3 Re5 6.Qf3 Be4 >>>> µ (-0.78) Depth: 11/28 00:00:10 10103kN >>>>1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kf1 Qxh2 4.Qxe4+ Bg6 5.Qe3 Rg1+ 6.Ke2 Qh5+ >>>> µ (-0.87) Depth: 12/30 00:00:15 14926kN >>>>1.gxf4-- >>>> µ (-1.19) Depth: 13/32 00:00:36 35237kN >>>>1.gxf4-- Qxf4 2.Qe3 Rg5+ 3.Kh1 Qg4 4.f4 exf3 5.Qf2 Qf4 6.Nxf3 Bd5 >>>> -+ (-1.69) Depth: 13/34 00:00:54 51719kN >>>>1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kh1 Qg4 4.f4 exf3 5.Qf2 Bd5 6.Rd1 Kg6 >>>> -+ (-1.78) Depth: 14/34 00:03:06 176328kN >>>>1.gxf4-- >>>> -+ (-2.09) Depth: 15/37 00:06:25 368131kN >>>>1.gxf4-- Qxf4 2.Kh1 Rg5 3.f3 Bh5 4.Rg2 exf3 5.Rxg5 f2 6.Qd3+ g6 >>>> -+ (-3.03) Depth: 15/38 00:10:17 576496kN >>>> >>>>(W, 15.08.2001) >>> >>>For comparison analysis of Deep Fritz on p800(64 Mbytes hash) >>>Deep Fritz searched for more than 10 hours at depth 17 but >>>it still looks at f3. >>> >>>Note that Deep fritz finds fail low for gxf4 in less than 10 minutes >>>and it probably means that with the best hardware Deep Fritz can >>>find f3 at tournament time control. >>> >>>New position >>>8/p4bpk/7p/3rq3/3Npp2/PPQ3P1/3R1PKP/8 w - - 0 1 >>> >>>Analysis by Deep Fritz: >>> >>>1.gxf4-- >>> ³ (-0.50) Depth: 1/3 00:00:00 >>>1.gxf4-- >>> ³ (-0.50) Depth: 1/3 00:00:00 >>>1.h3! >>> ³ (-0.28) Depth: 1/10 00:00:00 >>>1.h3! f3+ 2.Kg1 >>> = (-0.25) Depth: 1/10 00:00:00 >>>1.Kg1! >>> = (-0.19) Depth: 1/10 00:00:00 >>>1.Kg1! fxg3 >>> = (0.00) Depth: 1/10 00:00:00 >>>1.Kg1 e3 >>> ³ (-0.44) Depth: 2/14 00:00:00 >>>1.Rd1! >>> ³ (-0.34) Depth: 2/14 00:00:00 >>>1.Rd1! f3+ >>> ³ (-0.28) Depth: 2/14 00:00:00 >>>1.Rd1 Qh5 2.Qe1 f3+ 3.Kg1 >>> ³ (-0.41) Depth: 3/13 00:00:00 3kN >>>1.h3! >>> ³ (-0.37) Depth: 3/15 00:00:00 4kN >>>1.h3! f3+ >>> ³ (-0.34) Depth: 3/15 00:00:00 5kN >>>1.gxf4! >>> ³ (-0.31) Depth: 3/16 00:00:00 6kN >>>1.gxf4 Qxf4 2.Qe3 Qg4+ 3.Qg3 Qh5 >>> = (-0.25) Depth: 4/14 00:00:00 7kN >>>1.gxf4 Qxf4 2.Qe3 Qg4+ 3.Qg3 Qh5 4.Kg1 >>> ³ (-0.31) Depth: 5/15 00:00:00 20kN >>>1.gxf4 Qxf4 2.Qe3 Qg4+ 3.Qg3 Qh5 4.Kg1 a6 >>> ³ (-0.31) Depth: 6/17 00:00:00 40kN >>>1.gxf4 Qxf4 2.Qe3 Qg4+ 3.Qg3 Qh5 4.Qe3 Qg6+ 5.Qg3 Qf6 >>> ³ (-0.31) Depth: 7/19 00:00:00 83kN >>>1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kh1 Qg4 4.f4 exf3 5.Nxf3 Bd5 >>> ³ (-0.37) Depth: 8/23 00:00:00 152kN >>>1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kh1 Qg4 4.f4 exf3 5.Nxf3 Bd5 6.Rf2 >>> ³ (-0.34) Depth: 9/25 00:00:00 417kN >>>1.gxf4-- >>> ³ (-0.66) Depth: 10/28 00:00:01 801kN >>>1.gxf4-- Qxf4 2.Qe3 Rg5+ 3.Kf1 Qxh2 4.Qxe4+ Bg6 5.Qe3 Re5 6.Qf3 Be4 >>> µ (-0.91) Depth: 10/28 00:00:01 1262kN >>>1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kf1 Qxh2 4.Qxe4+ Bg6 5.Qe3 Re5 6.Qf3 Be4 >>> µ (-0.78) Depth: 11/30 00:00:13 9723kN >>>1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kf1 Qxh2 4.Qxe4+ Bg6 5.Qe3 Rg1+ 6.Ke2 Qh5+ >>> µ (-0.87) Depth: 12/30 00:00:20 14667kN >>>1.gxf4-- >>> µ (-1.19) Depth: 13/32 00:00:47 33715kN >>>1.gxf4-- Qxf4 2.Qe3 Rg5+ 3.Kh1 Qg4 4.f4 exf3 5.Qf2 Qf4 6.Nxf3 Bd5 >>> -+ (-1.69) Depth: 13/36 00:01:15 52680kN >>>1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kh1 Qg4 4.f4 exf3 5.Qf2 Bd5 6.Rd1 Kg6 >>> -+ (-1.78) Depth: 14/35 00:04:39 199152kN >>>1.gxf4-- >>> -+ (-2.09) Depth: 15/37 00:09:43 422090kN >>>1.gxf4-- Qxf4 2.Kh1 Rg5 3.f3 Bh5 4.Rd1 Rd5 5.Re1 e3 6.Qc2+ Bg6 >>> -+ (-2.78) Depth: 15/40 00:14:51 635762kN >>>1.f3! >>> -+ (-2.75) Depth: 15/40 00:32:40 1427443kN >>>1.f3! Bh5 2.g4 e3 3.Rd1 Bf7 4.b4 e2 5.Re1 Rxd4 >>> -+ (-2.66) Depth: 15/40 00:42:14 1841019kN >>>1.f3 Bh5 2.g4 e3 3.Rd1 Bf7 4.b4 e2 5.Re1 Rxd4 6.Qc2+ >>> -+ (-2.66) Depth: 16/42 00:59:57 2619083kN >>>1.f3-- >>> -+ (-2.97) Depth: 17/42 01:45:08 4599019kN >>> >>>(Blass, Tel-aviv 16.09.2001) >>> >>>Uri >> >>Look at the eval at the top. That's from Deep Fritz on my 2x1.4Ghz. >> >>In :54 seconds gxf4 fails low. And 10 minutes later, DF still hasn't found f3. >> >> >> >>Slate > >In this case your Deep Fritz is only slightly faster than >PIII800 >Your Deep Fritz on 2*1.4GH > >1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kh1 Qg4 4.f4 exf3 5.Qf2 Bd5 6.Rd1 Kg6 > -+ (-1.78) Depth: 14/34 00:03:06 176328kN > >For comparison my Deep Fritz on pIII800 > >1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kh1 Qg4 4.f4 exf3 5.Qf2 Bd5 6.Rd1 Kg6 > -+ (-1.78) Depth: 14/35 00:04:39 199152kN > >I do not understand the reason that your Deep Fritz cannot see >more than 1000 Knodes per second because I expect Deep Fritz on >2*P1.4GH to be more than twice faster than pIII800. >If deep Fritz used only one processor then it can explain >the result. > >Uri My Deep Fritz almost never takes 100% of both CPU's. Deep Shredder, Deep Junior, and Crafty all do however. DF usually takes 70% - 85% of both CPU's. And it just depends on the position. In Fritz 6, with Fritzmark, I am getting 1313k nps. According to that eval, and I am getting 948k nps. Let me try this again: Deep Fritz T29c - W,S 8/p4bpk/7p/3rq3/3Npp2/PPQ3P1/3R1PKP/8 w - - 0 1 Analysis by Deep Fritz T29c: 1.h3 f3+ 2.Kg1 = (-0.25) Depth: 1/10 00:00:00 1.h3 f3+ 2.Kg1 = (-0.25) Depth: 1/10 00:00:00 1.h3 f3+ 2.Kg1 = (-0.25) Depth: 1/10 00:00:00 1.h3 f3+ 2.Kg1 = (-0.25) Depth: 1/10 00:00:00 1.Kg1! = (-0.19) Depth: 1/10 00:00:00 1.Kg1! fxg3 = (0.00) Depth: 1/10 00:00:00 1.Kg1 e3 ³ (-0.44) Depth: 2/14 00:00:00 1.Rd1! ³ (-0.34) Depth: 2/14 00:00:00 1.Rd1! f3+ ³ (-0.28) Depth: 2/14 00:00:00 1.Rd1 Qh5 2.Qe1 f3+ 3.Kg1 ³ (-0.41) Depth: 3/14 00:00:00 3kN 1.h3! ³ (-0.37) Depth: 3/15 00:00:00 5kN 1.h3! f3+ ³ (-0.34) Depth: 3/15 00:00:00 5kN 1.gxf4! ³ (-0.31) Depth: 3/16 00:00:00 6kN 1.gxf4 Qxf4 2.Qe3 Qg4+ 3.Qg3 Qh5 = (-0.25) Depth: 4/12 00:00:00 7kN 1.gxf4 Qxf4 2.Qe3 Qg4+ 3.Qg3 Qh5 4.Kg1 ³ (-0.31) Depth: 5/15 00:00:00 20kN 1.gxf4 Qxf4 2.Qe3 Qg4+ 3.Qg3 Qh5 4.Kg1 a6 ³ (-0.31) Depth: 6/17 00:00:00 41kN 1.gxf4 Qxf4 2.Qe3 Qg4+ 3.Qg3 Qh5 4.Qe3 Qg6+ 5.Qg3 Qf6 ³ (-0.31) Depth: 7/19 00:00:00 78kN 1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kh1 Qg4 4.f4 exf3 5.Nxf3 Bd5 ³ (-0.37) Depth: 8/22 00:00:00 164kN 1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kh1 Qg4 4.f4 exf3 5.Nxf3 Bd5 6.Rf2 ³ (-0.34) Depth: 9/22 00:00:00 443kN 1.gxf4-- ³ (-0.66) Depth: 10/28 00:00:00 890kN 1.gxf4-- Qxf4 2.Qe3 Rg5+ 3.Kf1 Qxh2 4.Qxe4+ Bg6 5.Qe3 Re5 6.Qf3 Be4 µ (-0.91) Depth: 10/28 00:00:01 1378kN 1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kf1 Qxh2 4.Qxe4+ Bg6 5.Qe3 Re5 6.Qf3 Be4 µ (-0.78) Depth: 11/29 00:00:07 10610kN 1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kf1 Qxh2 4.Qxe4+ Bg6 5.Qe3 Rg1+ 6.Ke2 Qh5+ µ (-0.87) Depth: 12/30 00:00:10 15562kN 1.gxf4-- µ (-1.19) Depth: 13/34 00:00:24 36943kN 1.gxf4-- Qxf4 2.Qe3 Rg5+ 3.Kh1 Qg4 4.f4 exf3 5.Qf2 Qf4 6.Nxf3 Bd5 -+ (-1.69) Depth: 13/36 00:00:35 51319kN 1.gxf4 Qxf4 2.Qe3 Rg5+ 3.Kh1 Qg4 4.f4 exf3 5.Qf2 Bd5 6.Rd1 Kg6 -+ (-1.78) Depth: 14/34 00:02:05 187716kN 1.gxf4-- -+ (-2.09) Depth: 15/36 00:04:31 406187kN 1.gxf4-- Qxf4 2.Kf1 Qxh2 3.Qg3 e3 4.Qxe3 Re5 5.Qd3+ g6 6.Ne2 Qh1+ -+ (-3.37) Depth: 15/37 00:09:51 873035kN 1.h3! -+ (-3.34) Depth: 15/37 00:10:25 927017kN 1.Kg1! -+ (-3.31) Depth: 15/37 00:12:32 1121487kN 1.Kg1! e3 2.fxe3 fxe3 3.Rd3 e2 4.Kf2 Rxd4 5.Rxd4 e1Q+ 6.Qxe1 Qxd4+ -+ (-3.00) Depth: 15/37 00:13:26 1202691kN 1.f3! -+ (-2.97) Depth: 15/37 00:20:31 1864681kN (W, 15.08.2001) According to that eval, I am getting 1.5M. I guess I broke something. I was testing somethings earlier, and had it set to 1 thread, perhaps it didn't go back to 2. Oh well. Sorry about that. Slate
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.