TalkChess.com

George asked me to tell you he was able to sit at the computer long enough to start a match between Houdini 3 64 and Komodo 6 64. It is a "Best of 12" and is being run on 6 Cores. Here are the specifications he gave me to post:



Alienware AURORA_R4
Intel i7 w/6 True Cores
Factory overclocked
Fritz 11 gui
6 Cores each/64bit
256MB hash
Bases=NONE
Ponder_Learning=OFF
Perfect 2012b.ctg w/12-move limit
Time Control = 60'+15" (which he says is rather long and he thinks is best)
Match = "Best of 12" (6.5 points will win the match)



I have not had the pleasure of meeting Marco, who george seems to really respect. But when this match is concluded, Marco will be contacted by one of us (most likely me) to see which version he wants to run against the winner of this match. The winner of this match will play Marco's version choice- under exact same conditions, for what george has dubbed the "World Championship". george and I hope you will enjoy. (Updates will be given as often as possible during these matches. And Houdini-Komodo started approx. 30 minutes ago.)

*George said to make it very clear that Marco did NOT get a free pass to the championship match. Names were drawn and in that way Marco happened to receive the bye.



Thank you for your time,

Linda P. Speight

After 2 games have been completed:

Game 1- drawn (Houdini white)

Game 2- drawn



Thank you,

Linda P. Speight

Hi Linda,

I didn't read this forum for a long time, what's going on with George ? I'll send him/you an email.

Best wishes,

Gab

9 games have now been completed, and it is very interesting. Please note the change: After the first 2 draws I posted early this morning, and got in the bed with george- he told me he had been thinking and wanted to change it to the best of 24. Which means it will take 12.5 points to win the match. He said we had the cores and the time, and that it would make for a much better match w/ a more reliable outcome. So please make note of it. As I said, 7 more games have been played since Update 1 and here are the results after 9 games:




Alienware AURORA_R4
Intel i7 w/6 True Cores
Factory overclocked
Fritz 11 gui
6 Cores each/64bit
256MB hash
Bases=NONE
Ponder_Learning=OFF
Perfect 2012b.ctg w/12-move limit
Time Control = 60'+15" (which he says is rather long and he thinks is best)
Match= **"Best of 24" (12.5 points will win the match)




Total Score After 9 Games:


Komodo 6 64- 5.5

Houdini 3 64- 3.5


So there have been 7 draws and 2 wins by Komodo 6! As of yet, Houdini has not broken thru with a win. I know my result posting is not as neat as george's, but please bear with me. (I just copy and paste his specs)

Another update to follow sometime tomorrow.




george told me to tell each of you hello and he misses you guys-

Linda P. Speight

Thanks Linda !

This match is over, and seems almost like it just started. I asked george if he was surprised, and he said not much of anything surprises him any longer. He checked the games enough to see that there were no time losses, so posting of results continues. He told me that to many people- this result will be a "jaw-dropper". He says include him, because winning a match and winning it in this way are two entirely different things. So here we go:




Alienware AURORA_R4
Intel i7 w/6 True Cores
Factory overclocked
Fritz 11 gui
6 Cores each/64bit
256MB hash
Bases=NONE
Ponder_Learning=OFF
Perfect 2012b.ctg w/12-move limit
60'+15"
Match = "Best of 24"




A quote from george- "I could not have even begun to take this on if it were not for Linda. And she learns faster than I ever did. Thanks to you, sweetheart. (Take her out of my life, then go ahead and stick a fork in me- I'm done.) I checked all games and there were NO time losses. Only the few games when we were sleeping were not checked in task mgr. to be sure each engine was getting "99" when thinking, and neither ever took any core usage from the other. I would welcome anyone to go over each game with a fine tooth comb. If there are any errors I missed, I certainly would appreciate knowing. I am just not up to it.

Does this mean Komodo is stronger than Houdini? I don't look at it from that perspective. For my purposes, Komodo was strong enough to win this match and move on to the finals against Stockfish. Tho in this case, even with the very few games, it would be hard to make a case at this time control that Houdini "might" be better. But I will leave that to the "stat guys". They know more than me.

In a few days Linda will be contacting Marco for me and getting the version number he wants us to use in his shot for the title. And congratulations to Don and Larry- some fine work was done here. (Maybe the moral to the story is if you want to be No.1, don't let me beta test for you.)"


Hope you all enjoyed these updates-

Linda P. Speight

Incredible !
Congrats Komodo !
Chers,
Gab

geots wrote:This match is over, and seems almost like it just started. I asked george if he was surprised, and he said not much of anything surprises him any longer. He checked the games enough to see that there were no time losses, so posting of results continues. He told me that to many people- this result will be a "jaw-dropper". He says include him, because winning a match and winning it in this way are two entirely different things. So here we go:




Alienware AURORA_R4
Intel i7 w/6 True Cores
Factory overclocked
Fritz 11 gui
6 Cores each/64bit
256MB hash
Bases=NONE
Ponder_Learning=OFF
Perfect 2012b.ctg w/12-move limit
60'+15"
Match = "Best of 24"

Code: Select all

SP8-i7, Blitz:60'+15"  

Komodo 6 64-bit    +6/-0/=13   12.5/19
Houdini 3 x64      +0/-6/=13    6.5/19

A quote from george- "I could not have even begun to take this on if it were not for Linda. And she learns faster than I ever did. Thanks to you, sweetheart. (Take her out of my life, then go ahead and stick a fork in me- I'm done.) I checked all games and there were NO time losses. Only the few games when we were sleeping were not checked in task mgr. to be sure each engine was getting "99" when thinking, and neither ever took any core usage from the other. I would welcome anyone to go over each game with a fine tooth comb. If there are any errors I missed, I certainly would appreciate knowing. I am just not up to it.

Does this mean Komodo is stronger than Houdini? I don't look at it from that perspective. For my purposes, Komodo was strong enough to win this match and move on to the finals against Stockfish. Tho in this case, even with the very few games, it would be hard to make a case at this time control that Houdini "might" be better. But I will leave that to the "stat guys". They know more than me.

In a few days Linda will be contacting Marco for me and getting the version number he wants us to use in his shot for the title. And congratulations to Don and Larry- some fine work was done here. (Maybe the moral to the story is if you want to be No.1, don't let me beta test for you.)"


Hope you all enjoyed these updates-

Linda P. Speight

Thanks for running the match. It confirms nicely what my scaling study showed, that Houdini may be stronger at blitz, but beyond a certain reasonable level Komodo pulls ahead.

Linda, there are scarcely any words for a wife who would carry on an ailing man's hobby like this. This is a stalwart example of love to the very utmost.

Nelson Hernandez wrote:Linda, there are scarcely any words for a wife who would carry on an ailing man's hobby like this. This is a stalwart example of love to the very utmost.

/agree