27 Jan 07
Originally posted by FabianFnas"Normal" is changing. Most home computers purchased in the past year have dual-core processors.
I think Deep Fritz can take advantage of dual (multiple?) processors and Fritz cannot. Normal home computers have not dual (multiple?) processors so there are no differences between Deep Fritz for normal home computers.
Keep the Fritz 10 and be happy with it.
28 Jan 07
Originally posted by LennyBruceHow much?
I would bet that the difference between Fritz 10 run on a fast machine, (like 3.gig) and Deep Fritz run on multiple processors would be insignificant.
Fritz on my son's machine (dual core 2.8 GhZ AMD) runs circles around Fritz on my box (3.4 GhZ Pentium HT). Deep Fritz would do more to take advantage of the dual core.
28 Jan 07
1 edit
Originally posted by WulebgrI'm interested in your experiments.....please describe the results. Does Fritz dual machine beat Fritz single easily? Under what time controls, etc.?
How much?
Fritz on my son's machine (dual core 2.8 GhZ AMD) runs circles around Fritz on my box (3.4 GhZ Pentium HT). Deep Fritz would do more to take advantage of the dual core.
28 Jan 07
Originally posted by LennyBruceI did similar testing more than one year ago that I reported at http://www.chessexchange.com/forum/viewtopic.php?t=1549
I'm interested in your experiments.....please describe the results. Does Fritz dual machine beat Fritz single easily? Under what time controls, etc.?
My post this morning was a wager and prediction.
I don't have Deep Fritz, but I know that the single processor version runs faster on my son's dual core AMD than on my Pentium. I expect that his dual core would perform even better with Deep Fritz. I also suspect, though I haven't seen anyone discuss it, that Deep Fritz may perform better on my box than the single processor version due to the manner that Pentium's HT technology mimics dual core processing with a single processor (I think it fools Fritz into using only half the CPUs resources).
28 Jan 07
????....Yes, I know the search table/number of positions, etc., will be much different on a faster dual procesasor.
Are the moves any different? Have you played them against each other and seen any real difference in strength? Or are you just counting the search depth/number of positions and leaving it at that?
28 Jan 07
Originally posted by LennyBruceDoubtful. It is the same engine code only faster. Maybe have Deep Fritz against Rybka multi Processor?
????....Yes, I know the search table/number of positions, etc., will be much different on a faster dual procesasor.
Are the moves any different? Have you played them against each other and seen any real difference in strength? Or are you just counting the search depth/number of positions and leaving it at that?
It would however be 'faster' at solving 'white to move and win' positions that are 12 moves deep.
28 Jan 07
1 edit
Originally posted by giantrobotYes I agree, and that was my point. a chess program on a machine that's much faster will surely have millions more positions searched if you let it run long enough, but the search depth may only be 1/2 ply deeper, which usually doesn't change what the move will be.
Doubtful. It is the same engine code only faster. Maybe have Deep Fritz against Rybka multi Processor?
It would however be 'faster' at solving 'white to move and win' positions that are 12 moves deep.
I don't believe that equates to "running circles" around the prog on a single processor, which is why I inquired if wulbgr had actually played them against each other. I doubt there would be much, if any difference in playing strength and the result would be a lot of draws or verrrrrrry long games.