CPU/Mobo Hot CPU what can I do.

T

TheIcebreaker

Disciple
Well I own a Q8400 running at stock 2.6GHz on a G31ES2L on the stock Intel cooler.

I recently bought a HD4870 used one. This is a Palit Sonic edition with a dual slot cooler.

Now When I run games on it they run flawlessly till now but the CPU runs quite hot reaching a max of above 95 degrees.

I noticed the same and cleaned out the cooler, applied some Tuniq TX2 paste and thought that would be that.

TO my dismay the temps remain at the same levels which I am a little uncomfortable with.

I also tried to lower the VID voltage from the BIOS to a lowly 1.1V from 1.28V. No errors even when playing games at full load but the temps did not budge.

What more can I do? Will an aftermarket cooler solve the issue. If yes then where can I get one under or abt 1k. I would settle for a second hand one
if that would do the job.

Please help me guys. I am loving this card I got but am a little afraid that My CPU might give out if I keep it running at those temps. I dont wanna get a new machine for atleast 8 months.

Strangely enough the GPU stays under 60 degrees even at full load.


PS.: the CPU fan runs at 2600 RPM is that normal or is slower then normal.
 
I have Zebby bijli with the front LED fan which comes with the case replaced by a CM fan which runs at 2k RPM.
I have 90 mm fan at the back panel slot which the Bijli has.

My side panel fans gave out so currently I have the side panel removed.
 
TheIcebreaker said:
I took a reading using core temp too. They are the same as before.
Click to expand...

Reseat the heat-sink and reapply thermal paste using the following guide -- Benchmark Reviews: Performance Computer Hardware Tests / Thermal Paste and How To Use It / Guide - How TIM Works & How To Apply It Correctly | bit-tech.net

If the temperatures do not improve then you are facing other issues most chances a damaged temperature sensor too something wrong with the socket.

Also enter the BIOS and change fan speed setting to scale with higher temperatures.

Hope this solves your issue, Cheerio!
 
may be the heat from the gpu is building up in the cabinet.
but as the side panels are already removed it shouldn't be.
 
Well to get back to you guys after applying new paste and reseating the heat sink I am happy with the temps.
Runnig the AIDA64 stabilty test for 2 hrs no errors and the temps are 80 on core 0 and 70ish on other cores.

I have two completely different questions now:-

1. Why is temp of one core always like 10-12 degrees greater than others on full load. I have checked this with every possible utility out there. I now have like 12 diff temp checking programs on my machine. Either the sensor is broken on one core or it does have higher temp??? The difference drops to like 4 degree on absolute 0 load (44 on core 0 and 39-40 on others). How can I determine which one is the case here??

2. Now we come to overclocking. I set the FSB to 400 kept the core voltage on 1.225V(Still undervolted. the VID is 1.2875). The multiplier is still 8, so the proccy is buzzing along at 3.2 GHz which is not too much for a good C2Q. That makes RAM go at 960 Mhz @ 5-5-5-20 with the voltage on those at 2.0V. The system runs completely fine. Ran the same stability tests and no errors. Now comes the dilemma. I could not set my PCI-E freq to 100, System refuses to POST when I do that, so that is set to AUTO. Now after reading about it I found out that maybe the mobo makes the PCI-E freq in the same ratio to the FSB so that comes to 120 from the usual 100. What are the problems I can run into with that? My on board LAN which runs on PCI-E bus is underperforming slightly and the raised PCI-E freq maybe the most probable cause of it. I dont want the damned thing to die ofc. Is there something I can do to get the system to POST at PCI-E set to 100.

PS.: Mobo is G31-ES2L and the proccy is Q8400.

Please get back to me fast guys. :D
 
Back
Top