Still, there has been some progress for Houdini. I have run a couple of 9,000 game matches at 2'+2" time control between the current Houdini 3 DEV version and previous Houdini versions. The Houdini 3 DEV version scores 57.2% against Houdini 2.0 (+50 +/- 4 Elo), and 60.2% against Houdini 1.5 (+72 +/- 4 Elo).
As announced before, the Houdini 3 Release is scheduled for September/October, it may depend on the weather and the observing time with the new telescope ;).
> The Houdini 3 DEV version scores 57.2% against Houdini 2.0 (+50 +/- 4 Elo), and 60.2% against Houdini 1.5 (+72 +/- 4 Elo).
This is really quite impressive. Houdini 3 will be more than 110 Elo points stronger than Rybka 4.1! Kudos to Robert Houdart for this magnificent achievement!
All you seem to read from them (CPU) is that a) copied b and b)copied c) who copied most of his stuff from a) in the first place.
Anyway, lets get it first, I've seen more than enough rumours about new engines coming out....still waiting for all of them
I still don't think Sal's tuning of Houndini should be so casually dismissed.
> Well, I'm not holding my breath.
I am because I tend to find when engines are optimised, their performance seems signficantly better than that achieved by the test-houses. Deep Rybka 4 was certainly in that category.
I suspect we will have the same discussions when Deep Rybka 5 is released.
> If Houdini can survive , I would imagine so could Rybka from version to version
I wouldn't doubt the likelihood that Rybka 5 public release would be a commercial success unless it was very quickly found its performance was some way below expectation. As long as Vas can keep the interest going he should be OK but it is a tough market now particularly with so many good free engines available.
To a large degree, it depends what his ambitions are on the public side. Commercially it makes sense to have some diversity in case the cluster loses its attraction for whoever is paying for it.
All speculation of course so we shall have to wait and see.
> All speculation of course so we shall have to wait and see.
The release of an unfettered Rybka 5 that was able to take the lead once agan- would certainly breathe new life into this hobby and bring with it a sense of exhiliration - a shot in the arm that no other chess engine could give it. It would be no small coup for Vas, and would undeniably reinvigorate interest in the man and his work.
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill