WinTimerTester
http://www.mediafire.com/?io91a22pa6z8ntv

Bullu$: If you want to help fix this bug, can you run the above program for about 60 seconds when not overclocking and then run the same program when you are overclocking with MSI CCII and post a screen shot of each.

WinTimerTester uses two different methods to measure time in a Windows PC. In theory, these two methods should both run at the same speed so the long term ratio between these timers should be 1.0000:1. If one of these timers is not accurate after you overclock while in Windows, then that is the source of this problem.

I think the bug you've found is a Windows bug with the high precision timers. Windows can measure time very accurately but this "stopwatch" is sometimes based on the total CPU MHz. If you change your base clock speed by using SetFSB or a similar Windows program, the Windows high precision timer can go out to lunch. This bug can screw up a wide variety of programs besides RealTemp.

If you run those tests, I'll put fixing this on the things to do list. If you're willing to waste some CPU cycles, this problem can be fixed but the results might not be as steady as the RealTemp MHz usually are.

There haven't been enough complaints over the years so it hasn't been a high priority to fix this. I thought this problem was mostly fixed in Vista but then it came back in Windows 7. Have you upgraded to SP1 yet? I'm not sure if Microsoft has come up with a fix for this or not. It tends to be motherboard specific so not all boards have this problem. Windows is supposed to measure time based on a fixed timer that doesn't change speed when the FSB or base clock changes speed but that's not always the case.