Author: Axel Schumacher
Date: 15:32:20 09/29/05
Go up one level in this thread
On September 29, 2005 at 17:40:25, Albert Silver wrote: >On September 29, 2005 at 15:37:39, Axel Schumacher wrote: > >>On September 29, 2005 at 14:39:07, Lance Perkins wrote: >> >>>I just completed some code that gets around the 32MB-per-process limit of WinCE. >>>So, I took the engine for a spin, as guest at ICC. >>> >>>This is what MobileThinker (ThinkerWCEDLL hashsize=5) was up against: >>> >>>Statistics for Colossus(C) On for: 3:31 Idle: 1 >>> rating [need] win loss draw total best >>>Bullet 2834 [8] 5246 30 44 5320 3027 (07-Aug-2001) >>>Blitz 2984 [8] 7137 36 99 7272 3131 (26-Nov-2001) >>>Standard 2603 [6] 3311 104 115 3530 2726 (08-Jul-2001) >>> 1: Crafty v18.15 (2 cpus) >>> 2: Thanks to Dr. Robert Hyatt and Tim Mann for all their hard work. ASUS >>> NCCH-DL dual Nocona 2.8ghz-800FSB. >>> 3: Colossus is usually unattended and can not answer tells. >>> >>>Nothing really can be said from a single 15 0 game. Still, I myself am surprised >>>as there is at least a 40 to 1 difference in hardware speed (MobileThinker on >>>Axim X30 only manages 35KNPS). >>> >>>Not bad for a bad sportsman and a kitchen appliance. >>> >>>[Event unrated "184"] >>>[Site ""] >>>[Date "2005.09.29 01:44:07"] >>>[Round ""] >>>[White "guest7986"] >>>[Black "Colossus"] >>>[Result "1-0"] >>>1. d2d4 g8f6 2. g1f3 g7g6 3. c2c4 f8g7 4. b1c3 d7d6 5. e2e4 e8g8 6. f1e2 c7c5 7. >>>d4c5 d6c5 8. e1g1 b8c6 9. d1c2 c6d4 10. f3d4 c5d4 11. f1d1 d8b6 12. c3b5 e7e5 >>>13. c4c5 b6c6 14. b5d6 b7b6 15. b2b4 c8e6 16. f2f4 e5f4 17. c1f4 f6g4 18. d6f5 >>>g6f5 19. e4f5 d4d3 20. c2d3 g7a1 21. e2g4 f8d8 22. d3g3 a1g7 23. f4d6 e6f5 24. >>>g4f5 b6c5 25. b4c5 h7h6 26. f5c2 a8c8 27. c2b3 c6e4 28. d1e1 e4d4 29. g1h1 d4c3 >>>30. g3f2 c3f6 31. f2e2 f6b2 32. b3f7 g8h8 33. e2e6 b2f2 34. f7g6 f2f6 35. e6e4 >>>f6d4 36. e4b7 a7a5 37. b7e7 d4f2 38. g6d3 d8d7 39. e7e4 g7c3 40. e1f1 f2d4 41. >>>c5c6 d7g7 42. d6f8 g7c7 43. e4g6 d4h4 44. f8h6 c3e5 45. h2h3 a5a4 46. h6g5 c7g7 >>>47. g6g7 h8g7 48. g5h4 c8c6 49. h4g5 c6c3 50. f1d1 c3a3 51. d1d2 e5c3 52. g5e7 >>>c3d2 53. e7a3 g7f6 54. g2g4 f6e5 55. h1g2 e5d4 56. d3b5 d2g5 57. b5a4 d4c3 58. >>>g2f3 c3d4 59. a3b2 d4d3 >> >> >>Lance, >>I ran a test-tourney with MobileThinker against the Top competitors Hiarcs 9.5a, >>Fruit 2.1 (in PocketGransmaster) and Pocket Fritz 2.01 (Shredder 7); 10 minute >>games. >>MobileThinker won all three matches ! (each match 1,5 : 0,5 ). >>I think your program is now the strongest available for PDAs. >>I also tested on ICC about 10 games; it won all of them. >> >>Axel > >Come on.... How can you possibly draw any conclusions on its relative strength >against 3 opponents from 6 games? Of course, you can't. I made just a small comment. Unfortunately, it is soooo difficult to test, since you have to do everything manually. Certain is, that MobileThinker is damn good. I mean, you're not even talking about 6 games >against each, which would still be far too few, but 6 games _total_. > >Note that just because SMK wrote Pocket Fritz 2.01, that doesn't make it >Shredder 7. The engine in Pockt Fritz is Shredder 7.04 ported to PDA. >Unless you can show me that a series of positions or games where >they play the same moves at a given depth or node count. > >Aside from that, I do have one question regarding the strength of Fruit on the >PDA. It won in my test against Pocket Fritz both games. I think the transfer to PDA is not bad. For TogaII, I can test Toga only when using a second PDA, not by switching between programs (ponder off, of course), since CEBoard has a small bug (on my device), so that no other program can be accessed. So I postponed that test. Could you compare the speed of Fruit running on CEBoard with Pocket >Grandmaster? I just wonder whether they yield the same results. > >Finally, what about Toga II? > > Albert
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.