chess programs use plenty of floating point calculation as it decides the value of each position. Material value in professional chess software is generally not represented by integers as, say, a knight on the edge of the board is worth far less than a knight controlling key squares inside an opponent's ranks.
Running FritzMark wouldn't be as interesting:)
The match schedule says that they will play a certain amount of games with each AI on each platform, thus, with combined scores, eliminating the AI variable from the picture.
You're certainly right, they are 'non-classical', but not uncommon. Openings like that one are known as King's Indian Defenses and are part of a larger group of openings called Hypermodern.
This has been going on for some time...
on
Draw!
·
· Score: 1
The full install of Fritz 7 includes software to play chess on their Playchess.com servers. What's pertinent to this discussion is the 'Engine Room' area where you can load various chess-playing engines (like Fritz7, Shredder, Gambit Tiger, Junior et al...) and have your computer play someone else's computer. Generally, the person with the faster computer (and/or better engine, though most use Fritz7) has a much higher chance of winning. My computer (Athlon XP 2000+, GB of RAM [500mb for hash tables]) has played over 1200 games there (you can leave it unattended, so no I don't sit through all those games:P ) and is usually near the tops of the rankings. With tens of thousands of games being played daily there between engines, and considering the results support the 'faster computer wins, all engines being equal' theorem, statistical laws should help support the argument that chess IS a test of CPU might.
I'm certainly not advocating Anand's or Tom's or whoever should drop explicitly empirical benchmarks in favor of chess, but personally, doesn't a bit of direct machine competition throw a bit of fun (dare I say) into the whole 'this hardware is better than that hardware' (in this case, MP's and Xeons) benchmarking struggle?
Slashdot Mirror
Don't forget the canonical 'how many fps in [insert 3d game/benchmark here] would it get?'
chess programs use plenty of floating point calculation as it decides the value of each position. Material value in professional chess software is generally not represented by integers as, say, a knight on the edge of the board is worth far less than a knight controlling key squares inside an opponent's ranks. Running FritzMark wouldn't be as interesting :)
The match schedule says that they will play a certain amount of games with each AI on each platform, thus, with combined scores, eliminating the AI variable from the picture.
You're certainly right, they are 'non-classical', but not uncommon. Openings like that one are known as King's Indian Defenses and are part of a larger group of openings called Hypermodern.
The full install of Fritz 7 includes software to play chess on their Playchess.com servers. What's pertinent to this discussion is the 'Engine Room' area where you can load various chess-playing engines (like Fritz7, Shredder, Gambit Tiger, Junior et al...) and have your computer play someone else's computer. Generally, the person with the faster computer (and/or better engine, though most use Fritz7) has a much higher chance of winning. My computer (Athlon XP 2000+, GB of RAM [500mb for hash tables]) has played over 1200 games there (you can leave it unattended, so no I don't sit through all those games :P ) and is usually near the tops of the rankings. With tens of thousands of games being played daily there between engines, and considering the results support the 'faster computer wins, all engines being equal' theorem, statistical laws should help support the argument that chess IS a test of CPU might.
I'm certainly not advocating Anand's or Tom's or whoever should drop explicitly empirical benchmarks in favor of chess, but personally, doesn't a bit of direct machine competition throw a bit of fun (dare I say) into the whole 'this hardware is better than that hardware' (in this case, MP's and Xeons) benchmarking struggle?