Hi folks,

I've had my computer built for over a year now, and recently started crashing in games. It's been going on for a few weeks now.

I took the time to check my temperatures the other day, and found my Asus Geforce 7800GTX 256mb is running at 62 degrees celcius on IDLE!

I put it under load for 10 mins in 3DMark2005, and it shot up to 91 degrees celcius! This sounds very wrong. I opened up my case to see if anything was obstructing the path, but it's clear. The fan is definitely spinning, and it's not rubbing against anything.

I remember when I first got it, it was usually around 45 degrees celcius idle. Any clue what could have caused this and how I could fix it?

The problem seems to have started after I installed Windows Vista... But I can't confirm that. Could my video card's fan settings have been changed? I've read about an ATI tool to crank up fan speeds, is there something similar for nVidia cards?

Thanks!
-Will

My card idles at 46°C and gets to around 58° under load. The default idle threshold is 115°C although thats pretty extreme. It soundls like you're running a bit hot.

Are you running areo glass in Vista? I'm guessing that since that is a 3D desktop that perhaps more of the GPU is being used to draw the screen, even at idle. Other than that, make sure the GPU fan is spinning at the correct speed and check to make sure the HSF isn't coming separated from the GPU.

I should have clarified, it's running those temperatures in Windows XP.

From visually looking at the card, it doesn't appear to be coming apart, but I imagine just the slightest bit would make a large difference. Is there any way to check the numerical GPU fan speed? Thanks for the reply.

I don't remember if coolbits has fan control, but as Prozac stated, rivatuner does.

Try getting the card out of the case to inspect if there is any crud inside of it. I have a 6800 that was overheating pretty badly too. I opened my case to give maintenance and looked at the card. I could see there was a lot of crap obstructing the fan, so I cleaned it. Temps went from idle: 65 to 49 and load: ~105 to ~68. Just my piece of advice, man.

Okay, thanks for the advice so far everyone. I'll try raising the fan speed with Riva tuner and see if that helps and physically inspect the card.

Those temps are alright and as long as the graphics card is not experiencing any instability then there's not much to worry about. However if you want the card to be cool anyways then get the Artic Cooling NV Silencer (rev3) for the best air cooling. The Zalman VF900 CU if you don't have enough slot space. You can also improve cooling by installing a side intake fan directing cool air to the graphics card cooler.

http://www.arctic-cooling.com/vga2.php?idx=40

http://www.zalman.co.kr/eng/produc [...] 2&code=013

Well, I am experiencing instability. I don't know if it's being caused by the graphics card, but after seeing the temperatures, that is my hunch.

The rest of my system is running quite cool, so I don't think it's poor case ventilation. My Athlon 3500+ runs 33 degrees celcius idle, and 49 degrees celcius under full load. I've been trying to find the source of my instability, so I ran SP2003 on my computer for 8 hours, with no errors.

After about an hour or two playing a 3D intensive game, it will cause a system lockup. The computer doesn't turn off, but I can't get control back. I haven't taken a temperature reading before a crash, but I can imagine they are approaching 100 degrees celcius.

I would get third party cooling if I thought it would help my instability. So do you guys think the high temps are causing my crashes?

Sounds like replacing your HSF might not be a bad idea. As chuck mentioned the VF900 is a great cooler. Especially if you are looking for something reasonably cheap that's not too large.

VF900: http://www.newegg.com/Product/Prod [...] 6835118001

Also, the best VGA cooler I have ever seen reviewed is the Thermalright HR-03:
http://www.newegg.com/Product/Prod [...] 835109134R

Check out the THG GPU cooler roundup:
http://www.tomshardware.com/2006/1 [...] page4.html

The HR-03 + a fan absolutely slays the competition (and if it keeps the X1900XTX cool, it will keep your GPU frosty). But, it looks like more of a pain to install. The VF900 is SUPER easy (I have done it on 7600GT) to install. So, there are some tradeoffs. For the 7800, the VF900 should be good enough, so I would roll with that + some AS5.

I just thought of this because it happened today: occasionally when I start up my computer, I get all kinds of graphical glitches in bios and the Windows loading screen. Sometimes it's short segments of randomly colored pixels all over the screen, other times it looks like all the text/images are vibrating.

The thing is though, once I get out of the windows loading screen, everything is fine. Could this be due to a large temperature spike in the beginning because my HSF isn't working correctly? Or could my card itself be dying?

dean7: Thanks for the recommendations. I'm seriously considering getting a new HSF for my card. I want to get an HSF that will allow me enoughs space to do SLI later if I choose to upgrade. I was looking at the Arctic Cooling NV Silencer 5 that Chuck recommended because it pumps the air out the back, which would be more optimal for SLI I think (rather than blowing the warm air down on to the other card). It's also on sale for $19.99 on Newegg.

I just got a chance to physically inspect the card today. It looked fine from the outside, but when I peaked in through the plastic cover over the heatsink, I saw a big barrier of dust between the heatsink and fan. I cleaned it up and now I'm back to Idling at 47-50 degrees celcius! Under full-load, I didn't go above 75 degrees celcius, so I'm happy. Hopefully this solves my lockups.

Just wanted to give everyone the update. Thanks again!

I know this topic is quite old, but I'd like to discuss MY issue on this thread as well (google'd). My BGF 7800GTX OC idles at >60 C, as we speak it's at 62 C. I'm running Vista (non-Aero; basic), and the card is enclosed within a very well ventilated case (http://www.thermaltake.com/product [...] 003bws.asp). I am also using the latest nVidia driver set (175.16), but have attempted reverting back to previous driver versions to eliminate potential software issues. Aside from the obvious (changing the "stock" HS/F) I wanted to see if I couldn't figure this one out. Whenever I run any level of D3D/OGL the PC will reboot itself; no freeze/bsod, just a simple reboot/power cycle. Also, I have removed the card's HS/F and memory HS, cleaned it accordingly, and placed it back on. It really wasn't dirty to be begin with as I clean my PC periodically, but just a note. Any clues? This issue has been apparent for about a week and a half, now. I had no issues previous, that I can remember. I'm pretty astute and knowledgeable in this field, but I feel I could use a helping hand(s). Thanks, all.

Edit: I should have noted that the reboot is not instant; it is "random" -- essentially so long as it has not overheated. I have seen the core temp reach as high as >95 C before rebooting. Although much of the time, It does not reach that high before rebooting the system.

Anybody?

Once you take off the HS cooler make sure to put on some thermal paste