Author: Daniel Shawul
Date: 23:51:13 11/19/04
Go up one level in this thread
On November 19, 2004 at 23:15:44, Daniel Shawul wrote:
>On November 19, 2004 at 11:28:34, Anthony Cozzie wrote:
>
>>Well, I also did a pretty complicated algorithm, and I designed my own. If you
>>borrowed someone else's things might go easier for you, although I think that
>>takes a lot of the fun out of it.
>
> Did you create your complicated algorithm ,after sitting back and thinking?
>Well if that is the case you are more Genius! What i did was make something
>which searches in parallel, and then something which can make splits at points.
>Up to this point there is hardly any alogorithm,you just have to have good MT
>programming background. Then my own alogorithm of splitting at all points. And
>then i tried scott's
> - split at 1st move if some criteria is met
> - split at 3rd move if some criteria is met
>Now i can think of a better algorithm of my OWN as you did. But without having
>something to work on , planning is ridiclous.
>You are certainly a better programmer than me and i am no where to compete with
>you. But don't try to frustrate others just because you have one.
>
>>It is also possible you are simply more
>>talented than I am at this. However, I wouldn't say that it works until you
>>have run it at a dual for a week or so with no problems. I *thought* I had
I just want to let you know that it worked! Dann told me that it worked
this morning. But there is a resource contention problem problem which is
causing slowdown. So i am going to work on advanced things now...
mine
>>working after about two weeks of implementation :) I think maybe now I'm down
>>to only 1 bug that seems to only occur on ICC.
>
> Fabien or Tord may be good candidates for you if only they want to make their
>engine parallel.
>
>>
>>The next task for you is to measure the speedup you get.
>
> speed up? are you kidding? I just sent my version to Dann to test
>if it works at all(with 0.1% probability). Anyway thanks for asking!
>
>> BTW, I do have to take
>>back my comment about ABDADA. After talking with some people, it seems like
>>ABDADA gives a _really_ poor speedup (<1.3 or so, which is barely worth doing).
>>Of course, Scott's algorithm probably works better. Scott, what kind of
> Anthony READ what i wrote above!
>daniel
>
>speedup
>>do you get?
>>
>>anthony
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.