HI guys, i need one little help. I have nvidia gts450, in this week my pc start have freezzes. I make any tests and search the problem, this sensor of temperature is wrong. Have any program for disable temperature shutdown ? Or change the value ? Sorry for my bad english.
64C Idle, see movies on youtube on 80C Run game this system crash in 1minute. I have laser temperature after crash 100C in the system my laser says 50c.
If i understood correctly your GTS-450 is overheating.The thermal paste in your gpu must have dried up. If you can remove the heatsink cooler of the gpu and re-apply the thermal compound. Get some some decent thermal paste like Arctic MX-2 or MX-4 that is non conductive.
Make sure your GPU fan is still working. When is the last time you removed the GPU and cleaned out the dust?
The computer is clean, I do it frequently. I replaced the thermal paste and also not resolved. I did update the bios and also not resolved.
None of your friends can say is given to remove the security system that makes the plate off automatically?
^^^ ??? While gaming did you tried to touch with bare hands heatsink of that GTS 450? Is it hot or just warm enough but you still could keep your fingers on it? Just for test purposes: open rig completely, find some desktop fan and make him blow air to that GTS450. Computer still freezes after that? Are you 100% sure it's GTS450 overheating problem? Did you tested rig with some other PSU?
With the pc connected and msi afterburn / smart fan / nvidia inspector scoring 90 degrees. Step leisure across the board, especially in memories and on the opposite side of where the center of the processor and not get higher than 55 degrees temperatures.
Crazy thanks for your help, Yes, I can put the hand on the plate and quietly hold, feel only a slight increase in temperature, as if it had actually 100 degrees would be impossible to do this.
I have no choice, the card is discarded as if I open any 3D application it shuts down the computer. I've checked everything that was possible can only be a problem in the sensor. But if I do not think not be losing anything.
I did a test with Furmark program and recorded a log to send to you. more i can't send links :/ in 5sec my gpu temp jump 54cº for 100. And my system shutdown
This problem start have two weeks. My computer have two years without problem. Cooler is OK, fan is OK thermal paste is ok
Temperature is usually measured by resistor (less accurate), capacitor (more accurate), Diode or transistor (most accurate). They are usually located on GPU plate. Quick Google search did not gave me image where it is located :/ As for shutdown, it is internal stuff controlled by card's vBIOS. And I have no clue if there is tool capable to change this value, but I am sure it can be changed if found. Maybe someone, who remembers all nVidia vBIOS editors, may help.
Hi Brenio, for it to jump from 54°C to 100°C the GPU must not be contacting the heatsink properly. Have you tried re-seating the heatsink?
I've downloaded some tools and editors bios, but none that have this feature. Unfortunately, I also have searched on google about the location of the sensor but could not find too :/
My cooler was with noise, it started the problem, to suspicious that the board came to warm up for the inefficiency of the cooler and ruined the sensor. I replaced the cooler and the defect continued, changed the thermal paste and the problem persists. Currently I am using a cooler attached directly at the source for the same stay at full speed. My fan is bolted not believe there is a way in which she does not seated firmly.