totally bogus freq reading

[NoBugsOnMe]

Intel i5 on an MSI p55-53 board

I've set my BIOS to not use any eist or any variance.

Everest reports a stable clock , but CPUID is all over the place reporting 9X and 20X multipliers every two seconds

Tried the alt method and its even worse.

[The Coolest]

The upcoming version should fix frequency detection.

[NoBugsOnMe]

ok, thanks. Is it hard to do correctly?

[GigaByte]

My i7 920 D0 also has wrong freq.

[jarablue]

Mine has 1. wrong frequency 2. wrong stepping version 3. wrong temps.

Was this the april fools release?

Good program...uh otherwise.

[HallowedSoul]

here is the multiplier problem:

[The Coolest]

There are two different things:
1. There is a problem where you simply see completely insane numbers as the frequency, that is a bug.
2. The "low" clock speed. Here's the explanation why you see that:

Core Temp's new algorithm basically measure the CPU's active clock cycles, and calculates the frequency based on that. Even if you disable speedstep, when the CPU is idle it skips clock cycles to save on power. Since this is the proper way to read clock speed for the Core i* series CPUs I thought I'd implement it for the C2D chips as well since the same code just works on these chips too.
I may have to revert back to the old way of speed detection for those processors, I'm still thinking about how to proceed.
Try to run Prime or superpi, you'll see it clock right up.

[bjjones]

already reported bug, just thought I'd let you know it has the same issue on an i860. Went back to 99.5 and it's correct again, 99.6 shows very low multipliers and freq that bounce all over. Temp part seems to look about the same on both versions. Thanks for the handy little pgm - BJ

[dbasinger]

Hate to add more of the obvious, but ditto on the insane

CoreTemp0996.jpg

[The Coolest]

Please see this post.

[The Coolest]

I have a new beta up, please give it a go:
http://www.alcpu.com/forums/viewtopic.php?p=3426#p3426