>
>> >> Provide data from Intel and I'll concede defeat. I've personally run 166MHz
>> >> CPUs over clocked to 200MHz, and under clocked to 100.
>> >
>> >I'll check.. ;)
>> >
>> >That I'd be willing to swallow. However, I doubt that you ran the same CPU
>> >at both 200 *and* 100 MHz.
>>
>> >Here's an experiment you can try: run your 200MHz at 60-75MHz, and vice
>> >versa. I doubt you will have a single success.
>>
>> Big trouble...
>
>You didn't do the experiment I suggested:
>
>Run a 200MHz-rated CPU at 75/100MHz, and a 75/100MHz-rated CPU at 200MHz.
Oh phooey on you....the 166 without question uses the same dies as the 200.
I did do the test you mentioned...
>All the experiment you did was show a wide range. You haven't shown me
>yet that there aren't multiple distinct ranges (implying multiple dies).
>
>> First I dropped the cpu down to 100MHz, brought it in to OpenDOS (haven't
>> installed Linux yet) and ran Landmark Speed20 for about 2 minutes. No
>> problems. I then cranked it up to 200MHz, and same thing......I let it run
>> for about 15 mintutes...no problems. Right now it is still running at 75MHz
>> after 10 minutes....
>
>Dunno what Landmark is; is it similar to SPEC?
A mickey mouse performance loop popular back in the dark age of 386's. IE
the 200 got a rating of 1150, the 100 got 575. Not the best but all I
have on there right now.
>I'd be more impressed if
>it completed SPEC & kernel builds several times.
Well you will have to burn with anticipation because I'm done playing today.
Try your own hardware.
------------------------------------------------------------------
Elite MicroComputers 908-541-4214 http://www.psychosis.com/emc/