Computer Chess Club Archives

Search

Messages

Subject: Re: Number of positions in chess -- the rest of the story

Author: Ratko V Tomic

Date: 19:09:39 01/03/06

Go up one level in this thread

On January 03, 2006 at 20:35:09, Dann Corbit wrote:

>EC\EC32TEST.C 70 xttm=tmq2ns(ttm)/nsym;
>
>I guess that tmq2ns is a conversion from struct tm to nanoseconds from some base
>date.  Is that correct?

double tmq2ns(qword TimeDiff)  converts the counts of high resolution hardware
timer (which usually has about 1/4 microsecond per tick) to nanoseconds. The
averaging over tens or hundreds of thousands of code/decode loops, each loop
coding few thousands of symbols, yields the averages for coding time per symbol
which are repeatable down to 0.01 nanoseconds per symbol. For even better timing
accuracy, you can call process & thread priority functions thrHi() and thrLo()
in Qiutl.c.

• Re: Number of positions in chess -- the rest of the story Dann Corbit 19:14:41 01/03/06
  • Re: Number of positions in chess -- the rest of the story (code update) Ratko V Tomic 03:03:58 01/04/06
  • Re: Number of positions in chess -- the rest of the story Ratko V Tomic 21:29:21 01/03/06
  • Re: Number of positions in chess -- the rest of the story Ratko V Tomic 20:06:28 01/03/06

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.