so yesterday my graphic card decided to warm up

[- English -]

It was about 10:30 pm when i heard this weird beeping from my computer, and only when i was in game did it happen, so i immediately assumed it was my graphics card, as when it needed more graphics it made the noise. I checked my Nivdia monitor panel and it showed a whopping 80oC. Immediately i turned it off and opened the case up, the graphics card was really really hot...

I stuck my fan on it for 5mins to cool it down and turned it back on.. the card went up to the late 70s then i turned it off for the night hoping it would cool over night. This morning when I turned it on... heard this booom! nah ok thats a joke but my card temp went high again. I opened the side case and its sticking at a steady 65oC now. Im not sure why its decided to do this today, considering it has been running fine for the last 2 months in alot hotter weather than this current rain up north.

My graphics card is an asus 8800gts, the 640mb one. Any ideas why this could happen now? i have 2 fans in, + my processor fan and + the fan on the gfx card which i made sure is working... well its turning fast anyway. Maybe an installation of another fan maybe ideal, but why would this come on all of a sudden?

Ideas welcome tnx

 

[Gamah]

Checked the PCI-E voltage? Your motherboard could have been going screwy so check the volts running through the pci-e.

 

[- English -]

hows that done gamgams?

 

[Ballard]

70 +- 10 degrees is pretty standard under load for the 8800 family without aftermarket cooling. So things are not necesarily at the critical point just yet. Have you actually monitored the temperature before? If not it can always be a bit of a shock the first time you see the load temp of your GPU. They naturally are significantly hotter then system/CPU temperatures.

Having said that you are starting to get to the threshold where temperature starts to be very concerning. The first things i always do for errant temperatures are;
- Check the room temperature and case dynamics. Both make a big difference. In fact even summer does! on a hot day in summer you can expect system temps to be 10+ degrees hotter than your winter temps.
- If nothing there has changed then double check both power connectors on the card, if the lower voltage cable is not connected properly then the fan may not spin at full speed.
- Have you (or maybe a nerdy brother?) altered the clock/memory speed settings on the card? Very easy to do in most driver packages these days. If it is running that hot on normal settings then underclocking it a little is a possibility though not something I would normally do.
- If its still less than 3 months old I would think about exchanging it for a new one. If you got it from a enthusiast / friendly shop you should have no problems.
- Modern cards are much cooler when in dual slot configuration blowing the air out the back. A aftermarket cooler could be your last option if all else fails.


p.s. High unlikely the pci-E voltage has been changed but if it has you can easily check it in your bios. In 99% of bios revisions it should be set to 'auto' or 'normal'. PCI-e bus speed should also be 'auto' or 100mhz (also in the bios). PCI-E voltage actually refers to the voltage across the communication bus not the graphics card and really will has a very limited effect on temperature except in extreme circumstances, its is primarily of interest for stability issues not temperature issues.

 

[Ballard]

oh yeah forgot to mention, whenever I first get a GPU I always take the heatsink/fan off, clean the old thermal past off both the GPU die and heatsink, and then re-apply some AS-5 or something similar. Reattach and then run as normal. Sometimes certain brand names <cough>sapphire<cough> skimp on quality.

All you need is a phillips, some cloth and ofc the thermal paste.

 

[Gamah]

hows that done gamgams?

Can be done through bios (pc health?) or with some software that you can probably get through your motherboard manu's website/disk...although always take the readings with a pinch of salt as they are not super accurate.

But yes as someone said GPU's regularly get to 80C so its not that much of a biggie.

 

[SkarIronfist]

With the Nvidia cards, the temperature is monitored via the Ntune facility.

If you notice the temperature levels on the left hand side, are banded. So there is a defined safe level. You can get ntune to warn if it goes to high. Click on the GPU within the Temperature section when right clicking on the ntune panel.

The temperature will drop off when you are not playing 3d titles.

Ok I am just checking my temps for my 8800 GTX.

At the desktop its 61C.

69C = WoW at 1920x1280 very high detail 8 AA/4AS in Origmmar
DAOC 1920x1280 No AA or AS (DAOC doesn't really tax the card).

Weather atm is mild 20C, no Air Conditioning.

 

[SkarIronfist]

Double post.

 

[- English -]

Hmm its weird. Ive come to ignore the problem of the card... that seems ok, ok i think 80'c was a little high but its currently running in the 60s now.

So that didnt seem to be the problem.. Using SpeeDFan showed me core1 and core2 really hot... 70's again which i think was causing the beeping and the occasional blue screen of death + crash. I remember having this problem a few months back when i first bought my computer.

My processor seems to be the problem, it a little loose, what i mean is the clip broke off 1 of the corner pieces when i was originally installing it and i clipped the other 3 in, which seemed ok. However if it takes a knock it may come alittle loose and for some reason causes the temps to go really unstable. I should have probably requested a new processor fan although you know what its like getting a new toy.. you just wanna play with it

Would this be whats causing my problem?

another note: when i start my computer up, i get random freezes before all the programs are loaded on the desktop... if i open a few files/folders while its booting up, it seems to be fine which leads me to think its a memory problem, although my ram is good qual and is definetly installed correctly.

 

[Ballard]

Ok 70's is very high for some CPU's. Not too bad for AMD and older intel but if you are talking Core duo then you do have a problem.

My processor seems to be the problem, it a little loose, what i mean is the clip broke off 1 of the corner pieces when i was originally installing it and i clipped the other 3 in, which seemed ok. However if it takes a knock it may come alittle loose and for some reason causes the temps to go really unstable. I should have probably requested a new processor fan although you know what its like getting a new toy.. you just wanna play with it

Would this be whats causing my problem?

Yikes!!! you are putting way too much force on it when installing it if you break one of the clips. while it may not stop the chip from working it, as you say, could result in the CPU not sitting in the slot correctly. If the die is not 100% flush against the heatsink you will have major cooling issues no matter how good a cooler you get.

 

[Kuhan]

all the 8800 series are made to only use 40% of the fan speed, and yea 80 degree's is nothing for them sort of cards.

what i did was download RiverTuner V2.01 and turn up the fan to 100% it now runs at 60 - 70 degree's on most games which is proberly the safest option