toddwbucy at grandecom.net
Tue May 20 13:33:20 CDT 2008
I am not sure but it sounded like he was having problems with his
wintendo...(I could be wrong...just assuming on my part) in which case
lmsensors will not work. If this is the case and you are using a nvidia
card or board you might want to use their monitors.
On Tue, 2008-05-20 at 07:26 -0500, Daniel Givens wrote:
> On Tue, May 20, 2008 at 6:14 AM, Chris Lemire <good_bye300 at yahoo.com> wrote:
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA1
> > My computer has been overheating and immediately powering off when
> > playing a game on high or full settings for 20-45 minutes. I am not
> > certain whether it is the graphics card or the cpu. I have 3 case fans
> > in my computer besides the many others, but I think stock cooling must
> > not be enough. How can
> > I know which part caused the power down?
> Get the lm-sensors package (apt-get install lm-sensors or yum install
> lm_sensors) and run sensors-detect as root or equivalent. Once the
> proper kernel modules have been loaded, you can run sensors to get the
> current temps. Just look at your standard running temp. it's idle
> you're CPU is over 60C, you're running too hot. It's not unheard of to
> see a 20C jump in CPU temps during high load.
> If you want to isolate just the processor, check out this page for cpu burn-in:
> If you have a multicore processor, you'll need to spawn multiple
> instances of it equal to the number of cores you have to max out each
> core and get a good result. If everything goes without problem there,
> then it's probably your video. I know the ATI and Nvidia control panel
> software has a temperature gauge in it, but I don't know about it's
> recording abilities. You can probably find the temps in a raw format
> in /proc/acpi/ somewhere.
> Hope this helps
More information about the SATLUG