Hey people, I've been reading the site and forum for a couple of months now and I really need your help. After a couple of hours of my graphics card gets so hot that it shuts itself down as a precaution (extremely anoying when playing online i can tell you). It's only been doing that the last two months, but the card itself is nearly 2 years old, so age might have something to do with it. I am planning to get me a new computer in about 3 months orso and 'donate' this one to a relative, but in the meantime I need a cheap and quick solution. My card is a Geforce 8800 GTS 640mb. My CPU has a Scythe Mugen cooling it and my case is running with one 10cm fan (front near my hd's) and one 12cm fan (back). This means that everything in my case is cooled nicely beside my graphics card. The side of my case also has a big roster so air does flow in and out of there. Any of you have an idea how I can cool my card so that it doesn't shut itself down. I was thinking about placing another 12cm fan in front of the roster near the graphics card so that it gets cooled better but I don't know if that help a lot or if there is a better solution.
What temperatures are you getting to? It might not be the card itself... The only thing I would suggest you do is get a new card, or check that your cooler isn't faulty, as it sounds like you have a good amount of airflow anyway
I'm absolutly sure it is the card, yesterday when it shut down again i opened up the case, to check up on things, and nearly burned my finger when touching the card. I was getting back into the game because people were waiting for me so I didn't have time to check the temp, but i'm sure it above 80 celcius. Meh I don't want to spend much money on getting a new card as I'm saving for my new computer, but if that's the best solution I guess i have to.
akaska vortexx neo. although to take 2 hours of gaming first suggests a fault rather than overheating.
I've had two 8800GTX's, one went back because it was reaching 115 degrees and then crashing surprisingly! My present one still reaches 80 degrees tho...
According to Furmark my GPU temp is 78 Celsius, but it's still not as warm as it was last night. That akasa vortexx neo seems like a nice and cheap solution to my problem, I will look into later. Thanks for the help guys.
Maybe worth removing the card and checking that your HSF is properly mounted with thermal goo, and that it's not clogged full of dust and dirt. My 8800GT never gets over 45degrees, though I've not overclocked...
78c isn't hot for a gpu, while if there was a problem with the card cooling you'd be hitting 90-100c very quickly during furmark. If you've had the problem when doing graphically intensive programs mostly, i'd say it was a power problem. I also had a problem with a gpu similar to your situation in which I too thought it was overheating and tried futile attempts at cooling it. In the end I realised the psu was the problem and changed it. Ofcourse you could have a more than capable psu and i'm barking up the wrong tree.
+1 My card 8800GTX was reaching 110deg C in benchmarks, opened up the cooler and the cooler was clogged with dust. Suggest you take the black case off the cooler and clean the inside (dont just blow at it there's dust where you can't see it) it ran so hot it wouldn't run SETI ever again. Ran folding@home fine though.
Yeah PSU was my first guess, however I found a few sites where you can calculate your estimate power usage. So I entered all my hardware and found that during heavy load it should be around 300w, my 450w psu should be able to handle that I think. It was only after that that I changed my mind from power shortage to overheating, oh and I also noted that 78 was cooler then it was during a shutdown. If I have to take a guess it was somewhere between 85 and 95. I'm not saying that it's definitly my gpu cooling but it does sound likely. edit: yeah i'll check that first.
depends on the psu, some generic psu's can't output over 60ish% of their max claimed output and your psu may have become faulty. like i said heat sounds unlikely as 2 hours of gaming is a long time, most gfx heat problems show up in the 5-20 mins range.
Ok I'm not a great computer expert (seeing that I'm still unable to know what a Dremel is ), but wouldn't a psu problem also show up in a relative short time instead of a few hours of heavy load.
Your 450w could have low amperage on the rails to start with but generates a fault in which its output starts dropping and after 2 hours shuts down.
If it's two years old and you haven't been actively cleaning it in that time, then it's more than likely clogged up with dust which means the fan will be very inefficient. Based on Occam's razor, it's more likely to be dust than any problem with your PSU.
+1 2 years of caked up crap in the heatsink is your problem. Dust bunnies are evil! get the vacuum cleaner onto the exhaust and suck out all the crap. if you have a bagless vac, empty it before doing so, just so you can see how much crap comes out. may i strongly suggest doing this with the machine turned off!
Actually I clean my case 2 times a year and I cleaned it about a month ago, just after the problems started. Dust should no longer cause any problems, cause there is hardly any there ... and no I don't only clean the case itself but also all the hardware inside.
The 8800GTS isn't a particularly long card but at 9" it's still long enough to split most mid towers in two, if possible why not try improving the airflow in one half of the case that may be getting choked, what case do you have and which fan configurations? Remembering that hot air rises you should aim to get cool air in at the bottom and exhaust the hot air out of the top, and if you can you may want to reseat your GPU Heatsink with new TIM
