T410 GPU overheating?
Posted: Thu Sep 27, 2012 4:41 pm
Hello,
I'm running into problems with my T410 with i7-620 and NV3100 gpu.
My laptop started to shut itself down suddenly, without any notice.
I've checked the cpu temp and it looks fine. I've uninstalled any think pad software responsible for throttling the cpu/fan and switched the fan to be always on the max rpm, but the problem didn't go away.
I've installed linux, just to make sure it's not windows related.
When I run 'sensors' app, I get a list of the temperatures of my system and... what's weird.. the GPU reads 90-96 degrees under no load at all.
By no load at all I mean the terminal running [without X windows].
It goes from 90 to 98 just like that... There's nothing going on on the system and it's that high!
Recently I had to replace the cooling unit because of the fan failure of the original one. I bought the exact same model and fan works fine. It's clean.
What's weird, is that the GPU is not that hot when I touch the radiator/cooler/heat pipe. If it was somewhere around 90 degrees I would not be able to touch it.
Maybe it's not the GPU reading?
the sensor name is: nouveau-pci-0100
What's wrong with it?
I thought I didn't put enough or too much thermal paste when replacing the cooling unit, but I've reapplied it today, and there's a nice thin layer of the ceramique paste from Arctic Cooling.
I'm wondering if the new cooling unit is ok, beucase under heavy load, the heat pipes are much hotter near the cpu/gpu then a bit further away. The whole heat pipe technology should not allow that [from what I read at least]. Anyway... the temperature under no load is still weird.
Any ideas?
Thanks,
Krystian
EDIT:
one more thing.
I haven't used it for quite some time due to the cooling unit replacement. Now that I've started it seemed to work just fine. It went through SuperPI & Prime95 tests without any issues. Then I started downloading missing updates and after that it started to shut down. I'm not sure these are related, especially that the behavior is observed under linux. Anyway... now when put under stress it shuts down within 2 minutes.
I'm running into problems with my T410 with i7-620 and NV3100 gpu.
My laptop started to shut itself down suddenly, without any notice.
I've checked the cpu temp and it looks fine. I've uninstalled any think pad software responsible for throttling the cpu/fan and switched the fan to be always on the max rpm, but the problem didn't go away.
I've installed linux, just to make sure it's not windows related.
When I run 'sensors' app, I get a list of the temperatures of my system and... what's weird.. the GPU reads 90-96 degrees under no load at all.
By no load at all I mean the terminal running [without X windows].
It goes from 90 to 98 just like that... There's nothing going on on the system and it's that high!
Recently I had to replace the cooling unit because of the fan failure of the original one. I bought the exact same model and fan works fine. It's clean.
What's weird, is that the GPU is not that hot when I touch the radiator/cooler/heat pipe. If it was somewhere around 90 degrees I would not be able to touch it.
Maybe it's not the GPU reading?
the sensor name is: nouveau-pci-0100
What's wrong with it?
I thought I didn't put enough or too much thermal paste when replacing the cooling unit, but I've reapplied it today, and there's a nice thin layer of the ceramique paste from Arctic Cooling.
I'm wondering if the new cooling unit is ok, beucase under heavy load, the heat pipes are much hotter near the cpu/gpu then a bit further away. The whole heat pipe technology should not allow that [from what I read at least]. Anyway... the temperature under no load is still weird.
Any ideas?
Thanks,
Krystian
EDIT:
one more thing.
I haven't used it for quite some time due to the cooling unit replacement. Now that I've started it seemed to work just fine. It went through SuperPI & Prime95 tests without any issues. Then I started downloading missing updates and after that it started to shut down. I'm not sure these are related, especially that the behavior is observed under linux. Anyway... now when put under stress it shuts down within 2 minutes.