[CannedBullets]

Alright. Right now I set it to 1.42v in the bios (rest of my specs in my signature). It just failed at 1.41v. So if this fails at 1.42v should I try changing the CPU-NB from Auto to 1.25v?

Also, I noticed that in both HWMonitor and HWInfo64 it says that my CPU core temp is about the same as my socket temp. What's going on? Does that usually happen? When I was testing 4.2 GHz the highest the socket and core got were 61 C socket and 51 C core temp. Is it because of the fan over my VRMs?

Yeah I was able to get 4.2 GHz stable at 1.32v and CPU-NB set to auto on 10 hours of Prime95 blend. Its always core 5 that fails.

Yeah when I have CPU-NB on auto the voltage shows up as 1.18v in the bios also.

EDIT: Yeah I've gotten this at 4.5 GHz stable a while ago.

 

[hammong]

If core 5 is consistently the one that fails, you're probably at the end of your rope. Every CPU has it's limit, and it shows when you get to the point that one core consistently runs hotter and/or fails before the others. Increasing voltage won't help, and sometimes results in damage to that core.

Does your bios allow you to set individual multipliers for each core? I know on my X79 I can set the six physical cores independently.

Greg

 

[CannedBullets]

Quote:

Originally Posted by hammong

If core 5 is consistently the one that fails, you're probably at the end of your rope. Every CPU has it's limit, and it shows when you get to the point that one core consistently runs hotter and/or fails before the others. Increasing voltage won't help, and sometimes results in damage to that core.

Does your bios allow you to set individual multipliers for each core? I know on my X79 I can set the six physical cores independently.

Greg

No it doesn't have individual multipliers. But it was core 5 that kept failing also when I was stress testing 4.2 ghz. But even then I was able to get it stable at 1.32v.

I don't think I'm at the limit, I just need to decide whether or not I need to raise the CPU-NB voltage or Core voltage.

 

[logan]

I actually just barely got an FX6300, and I was also having a hard time hitting 4.5GHz and clearing prime runs.

I changed my CPU-NB to 1.24 (don't have .01 increments in my BIOS), and vcore to 1.4425 (only adjustable in .02 incriments). I also set my LLC to "extreme" and turned off all of my cool and quiet functions. That seemed to do the trick. Mine wasn't quite stable at 1. 425v at 4.5GHz. Core 5 had issues, and it failed with errors after 40 min or so.

However, I do have issues with cooling at certain ambient temps. It would fail prime after about 2-3 hrs, usually after I closed the window to the room for the night to go to bed. I tested my heat theory out by using a box fan on the side of the case, and bam, 8 hrs stable easy. I'm not sure yet if it was CPU temp (mobo has built in overheating protection, but doesn't say what temp it kicks in), or VRM temps. Either way, that fixed it, and now I at least know its a temp issue.

Good luck

.

 

[CannedBullets]

Its been over 3 hours. vcore is at 1.42. No problems on P95 blend yet. Core 6 is slow to finish tests but there's no issues with it yet.

 

[FlashFir]

I thought I recall Bulldozer and Vishera always failing on prime. Use IBT or something different.

Ex: http://www.overclock.net/t/1318995/official-fx-8320-fx-8350-vishera-owners-club/5600_100#post_18832421

 

[CannedBullets]

Quote:

Originally Posted by FlashFir

I thought I recall Bulldozer and Vishera always failing on prime. Use IBT or something different.

Ex: http://www.overclock.net/t/1318995/official-fx-8320-fx-8350-vishera-owners-club/5600_100#post_18832421

I think that's only if the version of Prime95 you're using is older than 27.7.

 

[gridlock33]

arctic freezer a-30 cpu cooler and a small fan blowing on the vrm area

1.45 vcore and 1.15 default on northbridge. ambient temp was 22c in my room

 

[cainy1991]

Hey all, try bumping your VDDA volt up the first step from stock...
Makes a very large difference in stability when you are in or
above the 4.5ghz area.