View Full Version : Zotac GF 9300-ITX overheating

02-03-10, 12:50 AM

I have a Zotac GF 9300-ITX with an E3200 and this system constantly shuts down.

First I thought it would be a memory error (3.6GB detected out of 4GB) and changed memory against
some I knew worked (pulled it out of my X38 system).

At one point it shut down faster than I could reach for the power button to restart it. Finally I made it into
the bios and entered the hardware monitor. CPU was idling at 40-43C and that seems ok. But the MCP
was at 76C and rising very fast. At 95C it shut off.

Ok, maybe I made an error in mounting the psu in a way that it sucked the air out of the case. This seems
to be counterproductive because the cpu fan tries to blow down and the psu tries to move the air in the
opposite direction. So I turned the psu around hoping this would cure the problem. Nada.

I'm trying to rma the board but wanted to ask if anyone else has this board (it's the new one with oc'ed
chipset - model GF9300-I-E) and has similar problems.


02-03-10, 01:06 AM
Sounds like the north bridge heat sink isn't contacting the chip properly.

You could probably fix it yourself to avoid having to RMA it.

I'd take the heat sink off (carefully...) and apply some new thermal paste. If there aren't any clips or push pins holding the heat sink on then its probably just attached with thermal adhesive, so make sure to buy some adhesive thermal paste before removing it.

02-13-10, 06:13 PM
Hi krokodil,
did you solve the temperature issue with your Zotac GeForce 9300-ITX?
Was it enough to provide better contact between the Nvidia chip and the heat sink?

I'm having to deal with the same problem, now.

Thank you,

02-15-10, 03:07 AM
Hi krokodil,
did you solve the temperature issue with your Zotac GeForce 9300-ITX?

I'm sorry to say that I've solved it another way :)
As this board was only 4 weeks old (well, I had it only for that time) I didn't want to void the warranty and took the struggle with my dealer to get him to RMA the board to Zotac. They finally have given up resistance and let me return the board.

Do you have a 1st gen (silver chipset heatsink) or 2nd gen (black chipset heatsink) board?
I have both and 1st gen boards don't seem to have problems (I can even turn on cpu smartfan to reduce noise and it works fine: cpu temp goes up 10C to 50C and chipset temp is 60C, nothing to worry about).
It's the 2nd gen boards (factory overclocked) that seem to have problems.

02-21-10, 10:53 AM
Hi everybody,
I'm writing just to inform that the power-off issue with my GF9300 (1st generation: silver heatsink) was solved by applying new thermal paste under the MCP heatsink.
I'm very happy! :D

Thank you and good evening,

02-22-10, 01:02 AM
i have a waterblock for the mcp ;) custom Danger Den block, UV blue


03-11-10, 09:23 PM
I also ran into this problem. I found this thread when searching into the problem and was about to attempt the thermal paste fix myself. In my case though, I found that the CPU heatsink was actually barely touching the MCP heatsink on the notch where the cord output is, and caused it to lift slightly off the chipset. I placed the CPU heatsink (stock intel/45nm) ideally for cord placement, so I had to rotate it a bit more so the CPU heatsink was no longer touching the MCP hs. This resolved the problem entirely. In case anybody else runs into this problem with this board, and runs across this thread, I would advise to check this first.

Hope this helps.