- Thread Starter
- #21
You mean BIOS 0503 isn't working for you as good as 404 was? There's a BETA 0501 and 0503, which one are you referring to exactly?
-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
P5K/Q6600 at 3.6GHz, odd problem
- Thread starter GFORCE100
- Start date
You mean BIOS 0503 isn't working for you as good as 404 was? There's a BETA 0501 and 0503, which one are you referring to exactly?
- Thread Starter
- #22
Been doing more testing.
I can run the following stable at even 4-4-4-12 1000MHz DDR2 and 1.46V Vcore:
- WPrime
- Cinebench 9.5
- 3DMark06
- 3DMark05
- PCMark05
When I got to OCCT I've found it can reboot. Still investigating.
I did find however that Speedfan reports my +12V rail as 11.19V
Very low indeed.
I can run the following stable at even 4-4-4-12 1000MHz DDR2 and 1.46V Vcore:
- WPrime
- Cinebench 9.5
- 3DMark06
- 3DMark05
- PCMark05
When I got to OCCT I've found it can reboot. Still investigating.
I did find however that Speedfan reports my +12V rail as 11.19V
Very low indeed.
- Joined
- May 15, 2006
Those are software readings and should never be trusted. I had a motherboard that report the 12v rail was receiving 16v. Know what I did?I did find however that Speedfan reports my +12V rail as 11.19V
I grabbed my trusty digital multimeter (DMM) and checked it for myself on a molex plug, it reported 12.04.
- Joined
- Nov 26, 2005
- Location
- Concord, NC
I believe Nandro has the Premium, so the BIOS is different. Last stable BIOS for the Premium is 0204, and the Beta is 0302.
- Joined
- Apr 21, 2007
- Location
- Yorktown, VA
Did you try what I told you? I have had this problem along with many other people and it is the NB voltage most of the times. This has nothing to do with PSU.
suggestion:
http://www.ocforums.com/showpost.php?p=5243528&postcount=33
success:
http://www.ocforums.com/showpost.php?p=5244188&postcount=37
suggestion:
http://www.ocforums.com/showpost.php?p=5243528&postcount=33
success:
http://www.ocforums.com/showpost.php?p=5244188&postcount=37
Last edited:
- Joined
- Aug 29, 2006
So what was the problem? Usually abrupt crashes are things like GTEngineer or Thideras had already said.
Another thing might be memory not being able to handle load, yadayada. Good idea to think it might be PSU, but I think you have a decent PSU to put enough juice. Like thideras said, software readings bleh. I have looked over you PSU before and from other people's testaments that it's fine for your rig and I wouldn't worry about the 12 rail.
One more thing to add is just to check the PCI-E power, power plug, BIOS voltage setting (south bridge) etc. Yes, I saw that 3dmark06 came through fine, but just to eliminate that possibility. I used to have a rig with just enough wiggle room that if for any reason I wobbled the case, the videocard would wiggle and system would crash. And also a crash from not enough sb juice.
But again, I belive the same with GT and thid, that it's the NB and NB cooling and/or memory. The type of crash you're describing, that's the most common issue.
Another thing might be memory not being able to handle load, yadayada. Good idea to think it might be PSU, but I think you have a decent PSU to put enough juice. Like thideras said, software readings bleh. I have looked over you PSU before and from other people's testaments that it's fine for your rig and I wouldn't worry about the 12 rail.
One more thing to add is just to check the PCI-E power, power plug, BIOS voltage setting (south bridge) etc. Yes, I saw that 3dmark06 came through fine, but just to eliminate that possibility. I used to have a rig with just enough wiggle room that if for any reason I wobbled the case, the videocard would wiggle and system would crash. And also a crash from not enough sb juice.
But again, I belive the same with GT and thid, that it's the NB and NB cooling and/or memory. The type of crash you're describing, that's the most common issue.
- Thread Starter
- #27
Well it's not the NB voltage, tried it at various settings, 1.40, 1.55, 1.70 and doesn't make a difference.
I need to keep on testing but the number of untested combinations in terms of BIOS settings is somewhat running dry.
So far only OCCT and Prime95 seem to be the ones that cause reboots. OCCT does so around the 9min mark.
It's not heat since the core temps never get to 80C and mostly sit around 69/72C. Overall CPU temp is usually 60C.
Setting the RAM at its rated 5-5-5-15 1000MHz 2.2V doesn't help.
I need to keep on testing but the number of untested combinations in terms of BIOS settings is somewhat running dry.
So far only OCCT and Prime95 seem to be the ones that cause reboots. OCCT does so around the 9min mark.
It's not heat since the core temps never get to 80C and mostly sit around 69/72C. Overall CPU temp is usually 60C.
Setting the RAM at its rated 5-5-5-15 1000MHz 2.2V doesn't help.
- Joined
- Nov 26, 2005
- Location
- Concord, NC
Have you tried messing with the other voltage settings?
- Joined
- Aug 29, 2006
With all that varying, did you check your NB temps? Maybe your NB or RAM is overheating?
- Thread Starter
- #31
Sadly I don't have one on me right now so can't say. I'm trying other options in the meantime.
I have a Cheiftec 850W PSU coming on Monday anyway so if that with 60A on +12V rails still causes reboots, we know it's not a power issue. This Enermax PSU will go in a 2nd system so the purchase was necessary sooner or later anyway.
The Cheiftec also have the new PCI-E 8 pin connectors which will come in handly come Nvidia 8900, 8950 time.
- Thread Starter
- #32
If by MB temps you mean whatever PC Probe/Everest reports as my MB temp then it never got above 43C if I remember correctly.
I touched the MB heatsink and while at 1.70V it does feel hot, it's not hot enough to burn my fingers with and a big difference from say the heat produced by a P5W DH i975X chipset which I also own.
The Mofset heatsink at the top doesn't even feel warm and while some say it may not be contacting the mofsets correctly, if I bend it slightly down (so to seat it flat on the mofsets themselves instead at a slight angle - due to the way its held down by Asus), and hold it, it still doesn't heat up.
Anyway, Prime95 will reboot almost instantly so there's not even time for heat to build up, sure the CPU heats up pretty quickly but the motherboard is being stressed for what, 30-50 seconds before Prime95 reboots.
I'm trying having the RAM run at SPD now instead of manual settings, will see how this goes.
Or maybe the 0503 BIOS is just bugged and I should try the 0404.
Nandro
Disabled
- Joined
- Mar 21, 2004
- Location
- Wherever I happen to be.
Prime95 has caused me nightmares with my q6600 quad. Although its totally orthos stable, it will sometimes still fail prime or fail on one core. or occasionally make it reboot as I believe it loads all the cores fully and quickly thus causing a vdroop enough to reboot.
- Thread Starter
- #34
The Vdroop on the P5K is very small or non-existant. It's 0.1V at most. The damper BIOS function comes in very useful.
It would be something on the board overheating real quick hence the reboot or error after X amount of time.
Nandro
Disabled
- Joined
- Mar 21, 2004
- Location
- Wherever I happen to be.
I'm using the P5K Premium watercooled. Just saying for whatever reason Prime 95 can cause weird things to happen. I think its like yanking something to hard and too fast. I can be stable on every other test but that at 4ghz.
- Thread Starter
- #36
I did touch my RAM and it can feel hot but I also tried running it at 800MHz 4-4-4-12 which made it run cooler, however not fix the reboot bug.
So far I can depending on the settings run OCCT for up to 16 minutes or so without it rebooting but at around this time (i.e. 16 minutes) it will crash (system continues to be stable in Windows, just OCCT says "Crashed"). So while the system doesn't reboot, there must be some rounding errors somewhere.
I'm trying the 8x450 = 3.6GHz approach now but Windows and Coretemp report the CPU speed as 4.05GHz. This would be fine except 8x450 = 3600 and not 4050. Where's the 9x multiplier kicking in is beyond me.
Nandro
Disabled
- Joined
- Mar 21, 2004
- Location
- Wherever I happen to be.
What ram are you using? I had major issues with my OCZ and am RMA'ing them now. Hopefully its not a compatability issue. Ballistix seem to work great.
- Thread Starter
- #38
Ballistix PC2-8000, the 1000MHz variety but they do 1200MHz at 5-5-5-15 just fine.
2x 1GB DIMM's.
With the 404 BIOS I ran Everest's stability test for over 3 hours without any problems.
Nandro
Disabled
- Joined
- Mar 21, 2004
- Location
- Wherever I happen to be.
All in all, an odd situation. As suggested, have you upped the voltage on the NB to 1.7 or so? You could piont a fan directly on the HS to cool it if necessary.
- Thread Starter
- #40
Yes I have tried 1.7V for the NB.
I need to re-test all I tried with this 404 BIOS now as I ran out of ideas with the 503 BIOS (which is BETA anyway).
Tomorrow being a Monday is not helping (Work and all) but if it's still flakey with BIOS 404 it should make itself known fairly quickly.
So far Everest stability test, PCMark05 and 3DMark06 all completed OK (no reboots, errors or being thrown back to the Windows desktop kind of thing).
Tomorrow I should also have the new Cheiftec 850W PSU to try out.
