Videocard overheating - Looking for a cost-effective solution!

darkiron112

Honorable
Jul 20, 2013
7
0
10,510
Hey everyone,

I'm a college student, 20, who's just moved to Pittsburgh and I'm on a tight budget.

I used to have an nVidia GeForce 8800 GTX that died on me, but I've upgraded to a 550TI. The problem is, it overheats. On almost every game I play.

I also had an older, much smaller monitor that I used to use before I upgraded to this much larger TV. Now, the increased pixels make it an even shorter period of time before games crap out on me.

I've gotten wise lately and done benchmarks and tests. I've downloaded and used FurMark to push the card to its limits and determined the problem is definitely heat. I run the 1920x1080 Burn-In Test and within 3 minutes, the card's heat steadily rises from a default 35-60'C depending on my activity, to a harmful 90'C whereupon the screen flashes black, the drivers crash, etc. The exact same symptoms as when the card breaks down as I play games.

I'm looking for a more effective heating solution than what I have. I have four fans-- One on the back, one on the CPU, one on the GPU, and one on the side panel.

Due to heating, I've tried removing the side panel to increase air flow but this doesn't help much at all.

I've also tried running the fans on 100% speed with Precision X and then testing the heat in FurMark, but it climbs-- slightly more slowly-- to its limit and crashes.

I've re-applied Arctic Silver to the card between the GPU and Heatsink. I've kept the fans clean. None of this matters, the card is simply not getting cooled rapidly enough.

If you take a look at my specs (I'm unfamiliar with Tom's Hardware, so I'll repost them at the bottom here even if I added them to my Computer field in my profile), you'll notice that I have a ATX Mini-Tower case and a certain Gigabyte motherboard. If you take a look at the motherboard photos online you may notice the PCIe slots at the BOTTOM of the motherboard.

I believe this is the culprit of the overheating-- the gap between the fans and the bottom of the case is only about 1"-2". It never occurred to me before, but now I'm wondering if this isn't the big culprit in why the fans can't cool the GPU in time.

If anyone has any suggestions, please help. I've looked briefly into liquid cooling, which I think would more than adequately solve my problem, except the apparatus is either expensive or nonexistent for my GPU.

Even if the answer is just to get a new and bigger case, as long as I know I'm making a purchase in the right direction. I don't trust my own beginners' luck in this matter, not to make mistakes and waste money.


Thanks for the help,
DarkIron112

--
EDIT: Forgot to add my hardware profile.

--
Case: ATX Mini-Tower
Mobo: Gigabyte GA-MA785G-UD3H
PSU: Antec Neo Eco 400W
CPU: AMD Athlon II X4 640
RAM: 5GB Dual-Channel DDR2
GPU: 1024MB GeForce GTX 550TI
VID: E390-A1 Vizio 39" LED TV
HDD: WD 320GB SATA - WD3200AAKS
HDD: SG 2TB SATA - ST2000DL004
DVD: LiteOn iHAS222 DVD R/W SATA
OS: Windows 7 (x64) // Ubuntu 13.04 (x64)
 
Solution
Not likely you have a case temp problem if you have failed the same way with the case open. Double check using a room fan blowing right into the open case and see if the card temps rise the same way.

Are you overclocking? Overvolting? Used evga precision X to change power limits? If so reset them. Double check your frequencies against stock frequencies. Is the card a factory-OC'd card? If so then turn down the frequency back to stock for your chip.

+1 redo the thermal paste on the video card heatsink.

If this fails then consider under-clocking the card. IF the card works underclocked you may just have a marginal card, or the card may have been hurt by prior over-volt or over-clock (google electromigration).
when you re appllied the thermalcompound to the gpu did you spread the paste?

you might want to redo it and try the 5 dots method evenly across the IHS of the gpu, if it does not have an IHS then just apply a small blob to the centre of the gou die. make sure to squarely place down the heatsink and give it a few slight twists and do not lift it at any point after its down then secure it.


also are the temps still bad with the case open to the air?
 
Not likely you have a case temp problem if you have failed the same way with the case open. Double check using a room fan blowing right into the open case and see if the card temps rise the same way.

Are you overclocking? Overvolting? Used evga precision X to change power limits? If so reset them. Double check your frequencies against stock frequencies. Is the card a factory-OC'd card? If so then turn down the frequency back to stock for your chip.

+1 redo the thermal paste on the video card heatsink.

If this fails then consider under-clocking the card. IF the card works underclocked you may just have a marginal card, or the card may have been hurt by prior over-volt or over-clock (google electromigration).
 
Solution