a*****g
发帖数: 19398 | 1 (转)Go Softwares Who Beat Human
MoGo is at the top of the list. One thing to mention, all victories using
big handicaps. I wonder how far the different is if pro player give only 2
or 3 stones, or even only a 0.5 point komi as a handicap.
In 2008, thanks to an efficient message-passing parallelization, MoGo won
one game (out of three) against Catalin Taranu, 5th Dan Pro, in 9x9 with
standard time settings (30 minutes per side). MoGo was running on a cluster
provided by "Bull" (32 nodes with 8 cores per node, 3 GHz); the machine was
down during one of the lost games. The results of this event were approved
by the French Federation of Go. MoGo also played a 19x19 Game against
Catalin Taranu and lost in spite of 9 stones handicap. However, MoGo was in
good position during most of the game, and lost due to a bad choice in a ko
situation at the end. The machine used for this event (the IAGO challenge,
organized by the company "Recitsproque") is a good one, but far from the top
level in industry.
On August 7, 2008, the computer program MoGo running on 25 nodes (800 cores,
4 cores per node with each core running at 4.7 GHz to produce 15 Teraflops)
[citation needed] of the Huygens cluster in Amsterdam beat professional Go
player Myungwan Kim (8p) in a nine stone handicap game on the 19x19 board on
the KGS Go Server. MoGo won by 1.5 points. Mr. Kim used around 13 minutes
of time while MoGo took around 55; however, he felt that using more time
would not have helped him win. In after-game commentary, Kim estimated the
playing strength of this machine as being in the range of 2–3 amateur dan.[
4] MyungWan and MoGo played a total of 4 games of varying handicaps and time
limits, each side winning two games. The game records are accessible on KGS
where MoGo played as MogoTitan. In a rematch on September 20, Kim won two
games giving MoGo nine stones.[5] On August 26, 2008, Mogo beat an Amateur
6d with five stones of handicap, this time running on 200 cores of the
Huygens cluster.
On September 4, 2008, the program Crazy Stone running on an 8-core personal
computer won against 30 year-old professional, Aoba Kaori (4p), receiving a
handicap of eight stones. The time control was 30 seconds per move. White
resigned after 185 moves. The game was played during the FIT2008 conference
in Japan.
In February 2009, MoGo won two 19x19 games against professional Go players
in the Taiwan Open 2009. With a 7-stones handicap the program defeated Zhou
Junxun (9p), and with a 6-stones handicap it defeated Li-Chen Chien (1p).
On February 14, 2009, Many Faces of Go running on a 32-core Xeon cluster
provided by Microsoft won against James Kerwin (1p) with a handicap of seven
stones. The game was played during the 2009 AAAS general meeting in Chicago.
On August 7, 2009, Many Faces of Go (version 12) resigned against Myungwan
Kim (8p) in a 7-stone handicap game.[10] Many Faces was playing on a 32 node
system provided by Microsoft. The "Man vs. Machine" event was part of the
2009 US Go Congress, which was held in Washington DC from August 1 to August
9.
On August 21 and 22, 2009, Zhou Junxun (9p) beat Many Faces of Go, MoGo, and
Zen in full-board 7-stone games, beat MoGo in an even 9×9 game, and won
one and lost one even 9×9 game against Fuego.
On July 20, 2010, MoGoTW won an even 9×9 game as white against Zhou Junxun
(9p).
On July 20, 2010, at the 2010 IEEE World Congress on Computational
Intelligence in Barcelona Spain computer program Zen played professional 4
dan Ping-Chiang Chou of Taiwan 19x19 Go. Yamato of Japan wrote Zen. Zen had
6 stone handicap. Each side had 45 minutes. Zen won the game.
On July 28, 2010, at the 2010 European Go Congress in Finland computer
program MogoTW played European professional 5 dan Catalin Taranu 19x19 Go.
MogoTW had a 7 stone handicap. The computer won. MogoTW is a joint project
between the MoGo team and a Taiwanese team.
Posted by Biondy at 1:49 PM
Labels: Go Software | a*****g
发帖数: 19398 | 2 职业5段让7子可以获胜
说明电脑的水平差不多 1D-2D 了吧
cluster
was
【在 a*****g 的大作中提到】
: (转)Go Softwares Who Beat Human
: MoGo is at the top of the list. One thing to mention, all victories using
: big handicaps. I wonder how far the different is if pro player give only 2
: or 3 stones, or even only a 0.5 point komi as a handicap.
: In 2008, thanks to an efficient message-passing parallelization, MoGo won
: one game (out of three) against Catalin Taranu, 5th Dan Pro, in 9x9 with
: standard time settings (30 minutes per side). MoGo was running on a cluster
: provided by "Bull" (32 nodes with 8 cores per node, 3 GHz); the machine was
: down during one of the lost games. The results of this event were approved
: by the French Federation of Go. MoGo also played a 19x19 Game against
|
|
