I twice run the HWbot wPrime v1.55 on the 1024M test with same results:

 

 

wPrime is telling me, that I cheating. What the hell...? I just use overclocked Pentium 90 to 125MHz and that it is. The machine is stable, it did SuperPi 32M anyway:

http://hwbot.org/submission/2971047_

 

...and I did not run any software over standard WinXP. In fact, I did not even touch the machine. And the computing time is not that great, after all. 15,75hours aren't anywhere near fast to get to the 94%

 

So, what could I do to produce P90 @ 125MHz wPrime score?

That can be down to instability, from what I know. Unlucky.

Okay, I try wPrime 1024 at lower FSB... but that will make the run even slower. In the long run, I will add some caps to help the stability at FSB 83 - I cannot add more Vcore (3.5V), so... maybe mod the VRM then. Oh, well - time will tell, I go for it.

That is the problem with time - transfer from 23.59 to 00.00 makes wprime crazy. So for long runs you either need to adjust system time during run to avoid this or start the tests that you are sure to take less than 24h at 00.xx system time

What the... :-O That makes sense, but... this is a serious bug. Dunno if to hope not confirmed or confirmed on this... but we will know in like 17h ...!

wPrime 1024M test finished w/o any troubles now!

 

 

A big thanks to TerraRaptor for help - w/o him I did not know, that "the problem with time - transfer from 23.59 to 00.00 makes wprime crazy" ... So I set the time to 0:01 and then I run wPrime 1024M test again, for the 3th time. This time no cheating complaining from wPrime!

Thanks, TerraRaptor! :celebration:

 

There is the score:

http://hwbot.org/submission/2973738_

 

A "bit" slower that I expected, but... time for optimizing Windows come

I didn't know that caused an issue either. We all learn something every day.

 

I've never run something that slow, I haven't got the patience.

Heheh, try some AMD K5 or P90...