Author: Christophe Theron
Date: 18:07:32 04/12/02
Go up one level in this thread
On April 12, 2002 at 13:48:38, Roy Eassa wrote:
>On April 12, 2002 at 13:33:27, Christophe Theron wrote:
>
>>On April 12, 2002 at 12:05:09, Roy Eassa wrote:
>>
>>>On April 11, 2002 at 23:20:26, Christophe Theron wrote:
>>>
>>>>I have noticed that a PalmIIIx running at 26MHz for example is only slightly
>>>>slower than my m505 running at 54MHz.
>>>
>>>
>>>Why would THAT be??
>>
>>
>>MHz means nothing. You should know.
>>
>
>
>Wait a second. I thought it meant _everything_ if you are using the same CPU
>(and other major components) in both cases.
>
>I am well aware that MHz does not mean much between different CPUs and/or
>different programs and certainly is NOT proportional to chess rating, but I
>thought it WAS proportional to benchmark speed under the circumstances listed.
>(You did not say, "only slightly weaker," you said "only slightly slower.")
There is no reliable relation between MHz and benchmark on the Palm because:
* there are different processors used. There is the DragonBall EZ in the
PalmIIIx and the DragonBall VZ in the m505. Wait states can be disabled with the
EZ, they are already disabled in the VZ. There are other settings like this that
you can play with when you overclock.
* the frequency generator of the DragonBall goes banana when you push it too
hard. You ask for 54MHz and it might as well give only 45MHz. AfterBurner on my
Palm says it is running at 54MHz, but I suspect it is running slower than that.
Bottom line: the MHz data is useless. The only reliable thing is the TigerMark,
provided that you check that your clock is running at the same pace as a real
clock (if it doesn't, you have overclocked too much and the frequency generator
is completely off).
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.