Author: Russell Reagan
Date: 16:31:47 12/31/03
Go up one level in this thread
On December 31, 2003 at 17:55:05, Ed Trice wrote: >Basically, all of the moves are precomputed indices into 80 bit attack arrays. >These are generated as fast as you can lookup data in a list. > >Given that was the case, I wanted to know how fast the lookup speeds were. That >number ended up being 140 million nodes per second on my test system. You have to be doing something to make your numbers look better than they are. 140 million moves per second on a 2GHz machine comes out to 14 cycles per move generated. Using 80-bit bitboards, a single bitwise operation will take quite a few cycles. That's just really hard to believe. You have all of these world class programmers, PhDs, people who write compilers for a living, and no one even comes close to your numbers. >Didn't I even email you the executeable? Not me. >I agree that something does not add up, but why do you run to the extreme end of >throwing out personal attacks? If something does not add up, why not try to >understand what is different in our metrics. Your numbers borderline on impossible, and much of what you write on your webpage gives the impression that you are either A) are intentionally trying to boost your numbers with unsound comparisons and straw men, or B) you really believe that your comparisons to Zillions of Games is an achievment. >I demonstrated what I was testing, explained it fully, sent you the test program >(I think it was you) and never heard back from you. Must have been someone else. I'm not sure it would mean much anyway without the full source code that I could compile myself and verify that your numbers are legit. >This is easy to reproduce, and it is not fiction. The Zillions program displays >its node count. Given Y different positions from different Gothic Chess >openings, and Z amount of time per position, you get a node/second average. > >As Zillions is the only other engine that plays Gothic Chess, what else could I >compare my own numbers with? > >When I took my nodes/second average and compared it to Zillions, the division >produced a number 30 point something. That means that Gothic Vortex is 30 times >as fast. I am not doubting that your program is 30 times faster than Zillions of Games. I'm saying that probably 95% of all programs will probably be at least 30 times faster than the Zillions of Games program, so it is nothing to write home about. Besides that, I don't buy the "Zillions of Games is the only thing I have to compare with" argument. There are literally hundreds of freely available chess programs on the internet, and dozens more that you can buy. The rules of gothic chess may be slightly different from chess, but the speed should be easily comparable. Since the chess pieces are a subset of the gothic chess pieces, it should be easy for you to compare the speeds of any chess program with your program. At the very least, you could assume a very small margin of error between the comparisons (a few percent). >I think this point has been touched upon already. If you can construct some >suite of tests, I would welcome the chance to compare data. > >By the way, on the latest Archbishop endgame I tested, Gothic Vortex averages >560,235 positions per second when it solves a mate in 16. > >Would you be interested in writing just a simple Archbishop and King move >generator for an 80 square board, then we can time each other's solutions for >positions we construct for these endgames? Playing ability of the program has nothing to do with this discussion. It's the 80-bit bitboard program that blows away every other PC chess program in existence that is in question. I don't see much point in writing a program for this. Yace is probably as fast or faster than what I'd come up with anyway, and you blow it away already. Besides, as I already stated, you should have no problem with comparing speeds with regular chess programs.
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.