Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Amazing results by Zappa. Does anyone know what are the innovations

Author: gerold daniels

Date: 04:53:28 11/14/05

Go up one level in this thread


On November 13, 2005 at 14:47:19, Zappa wrote:

>On November 13, 2005 at 14:12:28, Gian-Carlo Pascutto wrote:
>
>>On November 13, 2005 at 14:03:52, Andreas Guettinger wrote:
>>
>>>On November 13, 2005 at 13:44:44, Uri Blass wrote:
>>>
>>>>On November 13, 2005 at 13:34:27, Peter Wolff wrote:
>>>>
>>>>>On November 13, 2005 at 13:27:57, steve j s joyce wrote:
>>>>>
>>>>>>Hi all,
>>>>>>
>>>>>>   Zappa has clearly been having very strong results.  Does anyone know the
>>>>>>innovations that are being used, or is better hardware the key?
>>>>>>
>>>>>>Regards,
>>>>>>Steve
>>>>>
>>>>>
>>>>>Zappa does appear to be very, very strong, but exactly how strong it is will be
>>>>>seen only when it's released to the general public and can be tested under equal
>>>>>conditions (either as a single cpu version or a multi-processor version, playing
>>>>>other engines that are not handicapped by lack of strong HW).
>>>>>
>>>>>It seems that Zappa's release is imminent though :-)
>>>>>
>>>>>My prediction is that it will have to do much better than in the short events
>>>>>such as the WCCC or Leiden to show that's better than Deep Shredder on a fast
>>>>>dual or quad box, or even stronger than the new SMP Gandalf, for that matter.
>>>>>Not to mention the new kid on the block which is getting ready to go SMP,
>>>>>Fruit...
>>>>>
>>>>>In any case, we will soon be having lots of fun :-)
>>>>
>>>>I do not think that fruit is ready to go SMP
>>>>
>>>>I understand that Fabien plans to do it SMP only in spring 2006 and this mean
>>>>half year from now and today there are other things that are planned earlier
>>>>like backward analysis capability and weak levels.
>>>
>>>If Fabien has a decent SMP version already in Spring 2006 then he is very fast
>>>in implementing SMP because this means _major_ changes to the engine.
>>
>>Naah. It takes about 2-3 months of *fulltime* work.
>>
>>The question is if Fabien is working fulltime on it or not. Last time I talked
>>to him, he was!
>>
>>--
>>GCP
>
>I think it depends a lot on how aggressive you are:
>
>Global lock or not
>Iterative search or not
>Work stealing approach or something simpler like Aphid/ABDADA etc
>
>Something on the low end is probably a few weekends of work, and it will give
>pretty reasonable results on a small number of processors (say 1.5 / 2 or
>something).  Fabien is a good programmer, and I have no doubt he'll manage.
>
>anthony

Hello Anthony. Nice win by Zappa. A lot of talk about your program going comm.
Will you be selling it?
Will you use your own GUI and distribute It Yourself ?

Good luck,Gerold.



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.