Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: KNs in J9

Author: Uri Blass

Date: 18:00:34 12/18/04

Go up one level in this thread


On December 18, 2004 at 20:40:28, Harvey Williamson wrote:

>On December 18, 2004 at 18:21:08, Uri Blass wrote:
>
>>On December 18, 2004 at 17:47:24, Harvey Williamson wrote:
>>
>>>I left my computer on infinite analysis looking at the following position. I
>>>noticed after about an hour when Junior was searching at depth 18 that the kns
>>>had dropped from over 2000 to under 300!
>>>
>>>Why would that be?
>>>
>>>
>>>
>>>1. Nf3 d5 2. d4 c6 3. c4 e6 4. Qc2 Nf6 5. Bg5 dxc4 6. Qxc4 b5 7. Qc2 Bb7 8. e3
>>>Nbd7 9. Nbd2 a6 10. a4 Rc8 11. Be2 c5 12. d5 exd5 13. axb5 a5 14. O-O c4 15. e4
>>>dxe4 16. Nd4 *
>>
>>Most programs do not know to count nodes or to calculate nodes per second
>>correctly when the number of nodes is bigger than 2^32.
>>
>>If you give it to search for an hour then it is natural to get more than
>>2^32 nodes.
>>
>>Uri
>
>Thanks Uri!
>
>I have only ever noticed this with J9 but I guess from what you say all is
>probably ok?

I noticed that problem with most engines if I give them enough time to analyze
more than 2^32 nodes.

Junior is a fast searcher so searching 2^32 nodes may happen faster with it.

Uri



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.