Author: Vincent Diepeveen
Date: 05:52:11 04/11/03
Go up one level in this thread
On April 10, 2003 at 13:05:58, Jonas Bylund wrote: >On April 10, 2003 at 12:39:59, Christophe Theron wrote: > >>On April 10, 2003 at 12:02:05, Vincent Diepeveen wrote: >> >>>On April 10, 2003 at 11:37:50, Jonas Bylund wrote: >>> >>>>On April 10, 2003 at 10:27:57, Vincent Diepeveen wrote: >>>> >>>>>On April 10, 2003 at 10:11:21, Jonas Cohonas wrote: >>>>> >>>>>>On April 10, 2003 at 09:25:09, Vincent Diepeveen wrote: >>>>>> >>>>>>>On April 10, 2003 at 09:20:15, ERIQ wrote: >>>>>>> >>>>>>>>These are all great goals, but I like this order better. >>>>>>>> >>>>>>>> >>>>>>>>* A Linux/bsd version of Chess Tiger w/ great gui ie. Eboard or better. >>>>>>>>* A native ARM version of Chess Tiger for Palm >>>>>>>>* Chess Tiger 16 >>>>>>>>* ...and a few more projects that I prefer to keep secret >>>>>>>> >>>>>>>>basis for order is: >>>>>>>> >>>>>>>>1.I will have a freebsd system running in about two week hopefully (just waiting >>>>>>>>on hardware to arrive) >>>>>>>> >>>>>>>>2.As soon as I could buy a new sony palm I will. So I can win a game from time >>>>>>>>to time :) >>>>>>>> >>>>>>>>3.And ct16 should be last because ct15 is already too strong!! whether it's >>>>>>>>first or last on that silly list that everone likes, I can't beat it on a 486 >>>>>>>>comp. And yes I've tried shamlessly >>>>>>>> >>>>>>>>Just my two cents. >>>>>>>> >>>>>>>> Sign, >>>>>>>> Eriq >>>>>>> >>>>>>>a dual version of CT15 would kick more butt than ct16 or working at the unknown >>>>>>>freebsd. note that freebsd allows multiprocessing but multithreading at it i >>>>>>>cannot advice. >>>>>> >>>>>>I think a dual version of CTX would be great! maybe we should have a hands up >>>>>>here, to see if we can influence the order of things ;) (note: people with dual >>>>>>processor systems votes count double, ok maybe only 1.7 :) >>>>>> >>>>>>Jonas >>>>> >>>>>Some people are simply too lazy or have too much bugs in their software to get >>>>>stuff parallel well to work. >>>>> >>>>>It will be always like that. >>>> >>>>Well i don't know where that came from, but i am quite sure that an accomplished >>>>programmer like CT would have no problems making Tiger SMP. >>> >>>in which case he is just plain lazy now. >> >> >> >>Vincent I have explained that I have other priorities. > >I respect that you have other priorities, i can relate on a different level. > >>Why should I spend time on a task that I estimate is mostly a waste of time? > >You shouldn't if that's what you really think priorities aside, but i wonder how >you can call it a waste of time when 3 of the absolute top programs have SMP >implimented; Junior, Fritz and Shredder. Not to mention the other strong engines >out there which are SMP too... > >>Can you mention the benefits that has brought to you the multiprocesor version >>of Diep? > >More strenght (however minute that might be) > >>Have you won any tournament thanks to it? Has it improved your image, the >>perception of quality in the eye of your future customers? How much money (=fuel >>to continue improving your chess engine) have you made from it? > >This is not the point at all, we simply don't have, Diep running on a single >processor, results from serious competition (not that i know of) therefore there >is no precedence indicating what Diep/Vincent have gained from SMP. > >I disagree with Vincent on a lot of issues, but where i agree with him is: a >program that can run single, dual, quad or even 500 processors is more a program >of the future than one that only runs single. > >The only thing stopping me from buying Diep is the lack of WB/UCI support! my interface will load UCI engines. Winboard perhaps too though it is a too primitive standard to load it stable. It has very limited functionality and has major bugs. UCI lacks a few things too but is a very reliable protocol compared to winboard. >>How can you justify that it has not been an almost complete waste of time? > >How can you justify that compchess in general is not a complete waste of time, >well i am sure that what ignited a spark in you to write a chess program as >great as CT is not far away from the spark igniting Diep as we know it today, >the things we do as humans to create, is a fine line between waste of time and >fulfillment. Whatever we say i find CT15 a very good engine, let's be clear here. I am rooting for DIEP too at the worldchamps 2003. DIEP at 90 0 level at a dual K7 sometimes doesn't manage to finish a 10 ply search. In positions where eval sucks simply (that is especially opening) that matters a lot. In far middlegame only sometimes and in endgame IMHO nowadays engines search deep enough. But for a program with a lot of knowledge a deeper search will not hurt, that is my argument to search deeper. Another argument is the Brutus thing. It completely kills engines that it outsearches; when you have more knowledge AND outsearch in a non-dubious way your opponents (forward pruning other than nullmove that makes the length of positional lines seen shorter is very dubious thing) then a well tuned engine will completely kill the rest. I am sure not many will be happy to play DIEP in world champs ;) With exception of the game against brutus, diep could not complain about its play in general in world champs 2002 against the world top. It just lacked a few small things which a well tuned 500 processor will not have. DIEP played or a win against Fritz for many hours. a 10 ply search then played the move f6?? after which a few very deep searches of fritz did some very bad positional moves (i get impression he is shortening the search for positoinal lines a little which causes it to play positionally worse) after which diep came back and had a won rook endgame. there fritz played it better and therefore it became a draw. DIEP played for a win for many hours against Shredder. DIEP played for a win against Warp for a very long period of time. then a tactical lbunder was made (by accident i played with a version without extensions; i had tested it that morning and had overseen it). Against Junior, which had a won position against diep, it managed to come back, but because of stupidiness of this operator in big time trouble (less than 5 minutes left) i made mistakes adjusting time and a broken harddisk and i had not turned on my cache for some weird reason, which caused it to perform a 6 ply search at move 90 somewhere resulting in g3? a tactical blunder. In short 2 tactical blunders because of search depth. Perhaps i would have won from junior if it had not searched 9 ply. So SMP for DIEP definitely matters. It is true that i could improve my sequential search. But DIEP needs less nodes a ply than every other engine in the world if that engine is not doing forward pruning other than nullmove. Even if they do futility and nearly nothing in qsearch, diep still needs less nodes. So i am not so sure it is to blame myself that diep is not searching deep. The problem is simply that it is 20 times slower than other programs. Each new processor generation that difference gets less. It is like factor 10 now for latest K7 processors or so, also because other programs get more slow scanning chess code (slowest thing is scanning chess positions in eval, not evaluating all the patterns; this last thing gets done in a logarithmically way). So it would be a joke to say that DIEP is not profitting from SMP. Now imagine i might get an equal search depth in world champs 2003. >Jonas
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.