Author: Albert Silver
Date: 05:11:40 11/17/01
Go up one level in this thread
On November 16, 2001 at 19:19:14, Christophe Theron wrote:
>On November 16, 2001 at 17:57:56, Roy Eassa wrote:
>
>>On November 16, 2001 at 13:56:14, Roy Eassa wrote:
>>
>>>Agreed: having the strongest chess program for the PC will lend MS significant
>>>prestige.
>>>
>>>Anybody think they have a chance at accomplishing this without buying out Tiger
>>>or another of the current top engines?
>>
>>
>>For the record, I think the answer is No.
>
>
>
>I think the answer is yes if they are ready to spend several years on it.
>
>I do not underestimate the skills of the people working at Microsoft. But I
>think it will take them time if they want to start from scratch.
>
>If they are keen on this they will simply try to buy a strong existing chess
>engine, start selling it, then improve on it.
>
>
>
> Christophe
I think it would definitely take years if they did it from scratch, and the
investment would have to be huge. Think of it. Engine development nowadays
remind me very much of the 100 meter dash. We're still seeing improvements
appear, though not in the huge leaps and bounds department, but rather shaving
off a few hundredths of a second here and there. If you already run the 100
meters under 10 seconds, then those micro gains seem like leaps and bounds, but
otherwise, it is merely a few hundredths of a second. It's much harder to
measure if you try to analyze the greyer area of machine vs. man. The above
description was already limited to comp-comp testing for the most part. They
would (let's not forget this MS-Chess talk is speculation - but fine, let's
speculate) have yet to reach the the lower bounds of top strength, and then the
question of shaving off enough hundredths of a second (refining the Search Tree
and not forgetting the Eval) is much less easy to reach, even with enormous
funds. The Deep Blue guys were spared this in a way, not to besmirch their work,
because they could and did go beyond the hardware limitations that you PC engine
programmers are confined to. What are your thoughts of teamwork on an engine? I
know you and Ed have exchanged ideas and seen each other's work, but it still
continues to be a one-man job for the most part.
Albert
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.