Asus P8P67 rev3 shutting down under load.

Has anyone heard of any problems running FreeBSD with one of these motherboards? This is the "NON" overclockable version of the p8p67 so it is not over clocked. I got this board because I have no intention of over clocking in the future.

When I first installed FreeBSD, the system install seemed to go fine. When I started to compile Xorg I noticed that out of nowhere the system just shut down and rebooted. I tried several times to compile Xorg and the system kept shutting off. I finally got it to install but then when I tried to compile Gnome, I got the same thing.

I finally gave up and installed PCBSD so I could play around to see what was up. So far everything has been fine until I tried to play Enemy Territory. After a while of game play the PC just shut down again.

I also dual boot Windows 7 on this same box and have never had an issue with Windows. I have played ET for hours without a shut down. With the PC just shutting down it makes it hard to troubleshoot so I am left thinking that it must be the motherboard protecting the CPU. I have checked the CPU under load while compiling and it never seems to get above 57deg c so it is confusing me a bit.

Is there any reason that this particular Motherboard should have any extra stress running FreeBSD? I am using the amd64 install (now PCBSD).
 
What does the system do when it "shuts down"? Is it a controlled shutdown?

ACPI bugs could misread the temperature, but there are lots of possibilities.
 
How many watts is your power supply? Do your fans speed up during intensive tasks? Do you have adequate airflow? Maybe install a pci(e)-bracketed cooling slot fan? Power supply or graphics cards getting too hot?
 
My power supply is 850w. My case fans are all regulated manually and there are five of them. I am not sure how to tell if the cpu fan is speeding up in FreeBSD. :( It is very quiet and is the stock fan. Its hard for me to tell. The monitor shows that it is speeding up in windows.

The case stays very cool. I doubt very much that the video card is getting hot. I have a fan right above it that sucks air across it and I have a HUGE fan in the top of the case that sucks heat out.

Wblock, It does not appear to be a controlled shut down because it states that the drives were not properly dismounted when it boots back up.

Right now I am in the middle of stress testing the CPU in windows using Prime95 and the cpu temp shot up to 75 deg and stopped. It is hovering between 75 - 77 deg c and has been for the last 1/2 hour or so which seems like it is a little hot even during stress testing. I dont want to push it too much further. I wanted to see if the box shut itself down due to over heating the cpu. Then I could assume that I may have installed the cpu fan improperly. I am still wondering. The max temperature/TCASE is supposed to be 72.6°C so I am not sure if I am measuring the temperature accurately. I would think that the mobo would shut the PC down at some point??

Another thing that I have noticed. When I reboot the machine and go in to the bios, the cpu temperature shows to be steady at around 60 deg c but when I immediately boot back in to windows (or FreeBSD) the cpu shows to be hovering idle at about 35 - 40. I am not sure why it shows a different temp. I am using coretemp in FreeBSD and the ASUS AI Suite in Windows. They both display close numbers at idle.
 
Hi

I have been running the motherboard for nearly 3 months now. I have upgraded from source to 8.2-STABLE and have done numerous ports installations from source. For example: rebuilding all the documentation in all the languages is quite CPU intensive, and scrubbing 5 x 2TB disks is very IO intensive.

No problems with shutting down.

regards
Malan
 
Goose, are you using the stock cpu fan and heat sink?

I still havent come up with a solution to this issue unfortunately. At this point I would be afraid to attempt to compile a kernel or to update the system... :(
 
neurosis said:
Goose, are you using the stock cpu fan and heat sink?

I have Intel i5-2500 CPU @ 3.30GHz and 16 GB RAM using the stock fan. I have not yet managed to have all 4 cores 100% all the time, but quite often I have pushed 1 core 100% for several 10's of minutes. I have all 8 SATA ports with devices (6 x HDD, CDROM, SSD). The only change from the default BIOS is that I have activated AHCI.
 
Forgive my ignorance, but how are you monitoring your cpu loads? and if you are, temperature as well?

I am guessing that there is a temperature difference between the i5 and i7 cpu's. My fear is that I did not attach the cpu fan/heat sink properly but again, in windows it reaches 75deg C and the temp stops climbing.

You have a full system it sounds like... :)
 
neurosis said:
Has anyone heard of any problems running FreeBSD with one of these motherboards? This is the "NON" overclockable version of the p8p67 so it is not over clocked. I got this board because I have no intention of over clocking in the future.

What model is it excactly? As far as i know every p8p67 model have the ability to boost the cpu speed.
Did you checked the core speed while stress testing?

I overclocked a i5 2500 (without K) on a friends PC and got 3,9-4,1 GHz.
Temperature while running Prime95 was max 74deg C but not with the boxed heatsink.

===

If your heatsink isnt installed properly the fan should rotate pretty fast even all the time. So this could be an indicator.
Check out your BIOS maybe you accidently hit the "Performance" Button and your CPU is overclocked because of that.

After entering the BIOS it should look like this:

remote_image20101121-24357-t10co3-0.jpg
 
That looks exactly like my bios and I do not have it set to performance/turbo. My processor is not the K version. Maybe that is what I was thinking when I was speaking about over clocking.

This is my exact motherboard.

I have not tampered with the BIOS at all other than to change the boot sequence.

Oh yes, and the fan does slow down at idle. I cant remember the fan speed at idle however. I'll look again this afternoon when i get a chance.
 
neurosis said:
Forgive my ignorance, but how are you monitoring your cpu loads? and if you are, temperature as well?

Hi,

The easiest is to include the coretemp kernel mode via:
Code:
#kldload coretemp

Then you can monitor the temperature quite easily via:
Code:
# sysctl dev.cpu.0.temperature
dev.cpu.0.temperature: 35.0C

I compiled the whole gcc-4.6 compiler, this ran for more than 10 minutes with the CPU for several minutes at 90-100% - i.e. all 4 cores utilised and the max temperature reached was 64 degrees using the stock cooler. The below was created with Cacti which I use for some system monitoring:

figJo.png


regards
Malan
 
Back
Top