P-6860fx's GPU seems to overheat around 70C... why?

I've been having overheating problems occasionally with my laptop ever since I bought it. At first it was the CPU that would get up to the upper 70s or low 80s and then overheat, so I replaced the stock T5550 with the T9300. Now that temp rarely gets above 55C, but my laptop's still overheating. Now I've noticed, using CPUID HWMonitor, that my laptop seems to overheat when my GPU reaches 70C. When I say overheat, I mean the whole display shuts off, the sound gets stuck in a loop, but the multimedia buttons stay lit. I'll be playing a game and alt+tab out to check the temps, and every time I do that it'll be in the mid to upper 60s. I can't recall a time where I actually saw the GPU temp at 70 or above, which makes me think that's the point at which it overheats.

70C is hot, but it isn't overheating hot, is it? I've read threads on here where people have issues with their GPUs in the 6860fx reaching 80+ degrees, so I don't think it's normal for my laptop to be overheating at these temps. Based on those other threads, I'd think it would be safe for my GPU to go at least a little higher than 70C. Is there some option somewhere I could change to up the shutoff temperature (as long as that probably wouldn't harm my laptop?)

I've tried searching the forum numerous times, but I haven't been able to find another thread about my problem. Thanks in advance for any help, it's appreciated.

 

It seems that the problem lies in the drivers you are using. Try upgrading to the latest drivers for your machine and see if this still happens.

 

Have you cleaned the vents in the laptop recently? Dust almost always causes overheating.

As for drivers, gateways 195.62 driver has been doing very well for me on my 6860. I would say try that and use the driver upgrade sticky at the top of this forum to install it.

In regards to temperatures, my setup sounds very similar to yours. I replaced my cpu with a t9300 and saw a huge drop in cpu temp. During heavy gaming, I see the gpu at most get to 70c, typically around 65c. IIRC 105c is the danger point for the 8800 gts. Sounds as though your temperatures are rather normal.

As a side note, if your using HWmonitor, what does your TZS0 temperature display? This temp reading has been a source of headaches for me after I upgraded my cpu. Staying constantly 10 degrees C above TZS1. If heat is an issue, I would say look here as well.

 

I've used several different drivers: the official ones from gateway's site, nvidia's site, and some that I needed modded inf's for from laptopvideo2go.com. I don't think drivers are the problem, and I'm currently using the 195.62 driver. I also just reformatted my hard drive and installed Windows 7 (I had the RC until it started rebooting) so I don't have any old driver installations. I went straight to 195.62.

Also, I make sure to clean the vents at least once a month because they can get nasty pretty quickly. Sounds like my temps are roughly the same as yours, but my laptop still seems to overheat at times, way before the danger point (which sounds familiar from something I read).

My TZS0 is almost constantly 10 degrees C BELOW TZS1, neither get dangerously high though. Right now they're at 42C/52C, and they maxed out today at 52C/62C. I'm not actually sure what those temperatures are measuring though, I've got no idea what an ACPI is and a google search didn't help much. My CPU maxed out at 49C, and the GPU maxed out at 66C. I was playing GTA4 for a while today, which is why the temps maxed out where they did. I'm fairly certain it's my GPU that overheats though, because I can play Mass Effect 2 all day long on high graphics with antialiasing off, but if I turn it on then ME2 will eventually make my laptop overheat.

 

It's not overheating, just bad craftsmanship of the 8800mgts.

This probably won't fix your issue but it might increase the time between your black screens.

1. If you are overclocking, bring it down to stock clocks and never attempt it again.

2.Take your laptop apart and put AS5 on the gpu. Its a 1hr job and is very easy to do, just keep track of screws

3. Try to keep your gpu as cool as possible, so don't use drivers that bump up your temps.

4. ......

5. Profit...??

 

Thanks for the suggestions, I've never overclocked the gpu. On the contrary, I've actually tried to set up rules with nvidia's system tools to underclock my GPU a bit if it got too hot, but I could never get the rules to save.

Unfortunately, I can't risk disassembling my whole laptop since I need it for school for the next 2 months, but hopefully I could do that over the summer or something. If I remember right, getting access to the GPU basically requires taking the whole thing apart, and I'm sure a lot could go wrong with that.

Also, what do you mean bad craftsmanship? Do you mean it shouldn't get that hot, or it shouldn't kill my computer at only 70C?

 

It's not about how hot it gets but how many heat cycles it goes through. Microscopic cracks in solder that expand and contract with heat, when the gap gets wide enough the graphics card stops responding. This is why it only happens in some games and at variable temperatures, where the graphics card is stressed on and off constantly.

That reminds me, try not to bump or drop your laptop either.