Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Afterburner speed selection as a way to measure move response

Author: Christophe Theron

Date: 15:45:45 12/27/01

Go up one level in this thread


On December 27, 2001 at 14:27:10, Rafael Vasquez wrote:

>On December 27, 2001 at 12:38:48, Christophe Theron wrote:
>Christophe,
>
>I tested the following:
>
>speedselection        timetoreachP7atstartpos
>33                    2min06sec
>35                    2min
>40                    1min43sec
>41                    1min41sec
>42                    1min39sec
>43                    1min35sec
>44                    1min33sec
>45                    1min32sec
> |                        |
>54                    1min32sec



The data above shows that the right overclocking for your Palm is 45MHz.

If you set it to 54MHz, what happens is that the OS believes it is running at
54MHz when it is actually running at 45MHz.

So it will not compute any faster, but will just count the time incorrectly.

So for you it's definitely 45MHz the right setting in AfterBurner.



    Christophe



>And:
>
>Move response to 1.Na3 at game=30min
>
>speedselection        time            pos
>45                    28sec           10425
>46                    28sec           10425
>47                    28sec           10425
>48                    1min12sec       28180
> |                        |
>50                    1min12sec       28180
>54                    1min12sec       28180
>
>You can choose 47sec for fast moving or 50+ for deepsearching :)
>
>Regards,
>Rafael
>
>>You have just discovered the evil effect of overclocking too much. In your case
>>the Palm will not slow down, but will count the time incorrectly.
>>
>>That's why you must use the highest speed that does not distort Chess Tiger's
>>time measurement.
>>
>>
>>
>>    Christophe



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.