Are my GPU Temps Normal - using M1730 (8700M GT SLI)?

Core 1 is 69-73°C at idle
Core 2 is 71-74°C at idle.

I haven't stressed them, but the temps haven't changed much at all while playing Batman: Arkham Asylum at 1920x1200. They both got up to 77°C. I don't really play videogames and got this game just because this computer is apparently a gaming machine and it is actually quite fun!

I'm worried about temperatures because I've read about the incredible failure rates of the GPUs in this model and if its already getting wacky, I'll be contacting Dell (as this is a replacement for an older computer).

 

Yer thats about standard when playing games.

 

For the 8700m SLi, those are okay temperatures.

 

When should I become alarmed? I'd rather not wait till I have verticle lines running down my screen and BSOD's all over the place.

 

Well, to be honest, the official thershold is much higher. Seriously higher. What you need to do is make sure they breathe.

However, you probably still want an alarm level. I'd say that you need to start considering giving XPS support a call if the temps exceed 95. Some people here will tell you 90, but I'd say 95 degrees celsius.

 

Thank, Kade. Your guidance is much appreciated.

 

the cards are not failingf from heat. They are failing from going hot cold hot cold. See my sig' link.

If you want to give your card sthe best life only turn on the system when you want to game and then turn it off when finished. Don't let it sit there idling away doing nothing.

I've proved this. My system has been sitting here for 2 months since the least new GPU's and done next to 0 gaming just idling away and doing the usual bit fo editing surfing tv watcheing etc................. waiting ............waiting.........hey presto no cards now.

 

So you are saying that normal use of the computer, (i.e. leaving it on) is eventually going to kill these cards? Not simply the stress of gaming?

 

I have the 8800 gtx cards and yes that's what has happened for me and lots of others.

Have you seen this?

 

Yes, and that thread is what worries me. However, I wasn't aware that it was simply temperature cycling that killed the cards. i assumed it was from overheating.

I haven't read much about 8700M GT cards. It seems that the 8800 GTX's are the prime culprits...

 

This problem that Slammin's talking about, has been present in the G84 cores as well, which means 8600m and 8700m.

But Slammin', do elucidate me on one thing, considering that we've discussed this topic quite a few times in the M1730 lounge. Wouldn't turning on the machine just for gaming, subject the part to even worse thermal cycling? I mean, going from cool--completely off--to the highs of 70-80, and then going right back down to a powered-down system. I don't think there is a clean and easy way to escape this eventuality, because the goal is to avoid cycling temperatures, and that'll still happen--perhaps even worse at times--when you completely shut down the system and then switch it back on for GPU intensive work.

 

@ Kade

OK.

My thought is, when you switch it on the 25c to 79c isn't instantaneous.

When I turn on my temps are usually 45c (I live in the tropics). By the time the system has gone to desktop and my AV etc has settled down it'll be 65c and on it's way to 75c anyway. Either way you are going to have to turn it on once a day regardless.

So if you only use the 1730 for gaming then you may as well only turn it on when you are going to use the thing. Rather than have it sitting there cycling over and over.

 

Good point about not using the machine at all, if it is only for gaming, because yes, it can self-cycle when it is idle.

However, on the note of temperature change from off to on, while slow, still creates a very strong cycle. There will still be, perhaps not a very fast, but powerful and large change in termperature, and to be honest, I think that kind of puts us in a paradox.

Y'know, there's no clean way to insure the survival of these GPUs. Well, aside from perhaps proper BOIS control, or an overclocked X9000.

Anyway, I did some further research, because my sluggish memory kicked-in, and I realised that while the G84 and G92 had problems, they weren't the exact same problem. In fact, the 8700m issues were kind of resolved and while they heat up a lot, they don't tend to die as easily as a majority of 8800/9800ms. So to the topic-creator, I'd say that you need not worry much at this point. Hell, my old Toshiba X205 used to hit temps of up to 96 with overclocked GPUs and the SLi 8600s never died.

 

Kade, thanks again for the help. Out of curiosity, what was the difference in issue between the G84 and G92 chipsets?

 

G84s were also failing, but their issue was directly the matter of the cards overheating and dying. And by overheating, I am talking very high temperatures. Most of the new BIOS releases from Dell--especially A06-onwards--actually deal with this matter by kicking in maximum fan speed earlier, or so says the internet press. I buy their story on that count.

8800m/9800m cards real issue is thermal cycling - going very high and then going low, which compromises a bad solder job.

8700m/8600m cards issue was just that some of them were heating up too much and going bad, and that's where the newer M1730's BIOS thermal protocol is holding up nicely.