• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Undervolting & stability

T

thewhat

Member
I'm undervolting an X2 4000+ (AM2) on an Asrock ALiveNF6G-VSTA. Using RMClock.

I went down to 0.8 GHz and 0.8V. Run it stable with LinX and Prime. (Even at 0.775V.)

But here's what's interesting: While it's stable in those stress programs, when browsing or moving windows I get a BSOD or restart.
Can anyone explain what's going on?
If I had to guess it would have to do with graphics.. but why would undervolting (underclocking) the CPU affect graphics?
 
Last edited:
It could be that AMD's version of turbo is kicking in when you stress test. That may include upping the voltage alongside increasing the clockspeed to a level that is able to run stably. While you aren't doing anything particularly stressful (moving windows around, etc.), it remains at the settings that you specified and it turns out that .8 volts isn't enough to run at .8 ghz, causing the instability.

That's just a guess. Someone with more knowledge should chime in here as well.
 
AMD doesn't have turbo mode on any of their processors except Phenom II X6. How long did you test for? A couple hours or more is necessary to ensure stability.

Also try using K10STAT as another undervolter. Maybe RMClock adjusts offset instead of voltage for each power state.
 
SlitheryDee said:
It could be that AMD's version of turbo is kicking in when you stress test. That may include upping the voltage alongside increasing the clockspeed to a level that is able to run stably.
Click to expand...
Could be, but:
1. I saw no indication in CPU-Z of such voltage or frequency increase during stress testing
2. I now set it to 0.8V in BIOS and disabled CnQ and it works fine so far, both in stress testing and overall

But yeah, CnQ (or something similar) could have something to do with it. Or RMClock..
The only reason I had CnQ enabled was to make the RMClock scaling features work. (EDIT: now that it's at .8V and .8Ghz in BIOS it seems to be working even with CnQ off.. hmmm.. I'll do more testing)
I need frequency scaling based on CPU load.


richierich1212 said:
Don't use RMClock, try out PhenomMsrTweaker instead:

http://phenommsrtweake.sourceforge.net/
Click to expand...
I don't have a K10 CPU, the program wouldn't run.
 
Last edited:
You must log in or register to reply here.
Back
Top