Jump to content

Severe Temperature Issues.


70 replies to this topic

#61 TostitoBandito

    Member

  • PipPipPipPipPip
  • 172 posts
  • LocationWashington, USA

Posted 13 August 2013 - 11:34 AM

You aren't understanding, it's not the game. 100% on your GPU is 100% on your GPU. It makes no difference what application is doing it or how many frames it is rendering per second. 100% on your particular GPU and laptop is clearly very hot, which is bad. You should repaste that if you haven't. The heatsink might not have good contact from the factory. If the heatsink is even a little bit off level and not making solid contact with the processor die, temps can be astronomical. They also use garbage thermal paste, and you can do much better with quality compounds.

I run a 6990M in my laptop, which is very similar to your card. When I first got my machine a couple years ago peak GPU temps in FurMark were in the mid 80's. I repasted the GPU (and CPU), replaced and added some thermal pads on the GPU, and since then my peak temp is around 73C. The cooling in your particular laptop model can make a difference as well. Mine is very robust (Alienware M17x R3) with quite a large fan hooked up to multiple heat pipes off the heatsink, and same for the CPU on the other side.

Again, you can keep blaming the game if you want but it is not the problem. Playing demanding games on laptops at safe temperatures often requires additional custom work. You can't just take a laptop out of the box and expect it to be ready for this. That is why so many people have heat-related failures and issues with laptops that they game on. They don't take those steps to lower temps and improve cooling.

http://forum.noteboo...ide-w-pics.html
That is a good guide for repasting. It was made for my laptop line, but I'm guessing yours is probably not too different.

Edited by TostitoBandito, 13 August 2013 - 12:02 PM.


#62 efryt

    Member

  • Pip
  • Bad Company
  • 15 posts

Posted 13 August 2013 - 11:48 AM

I experience same thing. I suppose there is no limitation for graphic card in lobby, which probbaly should be. PGI please set maximum fps there to 60 and problem will be solved.

#63 1Sascha

    Member

  • PipPipPipPipPipPip
  • 401 posts
  • LocationMunich, Germany

Posted 13 August 2013 - 11:08 PM

Quote

I experience same thing. I suppose there is no limitation for graphic card in lobby


I'm going to repeat myself, but:

I really don't think this is MechLab-specific. I know it's the obvious culprit to suspect, since it uses the engine to render the Mech you're working on. But like I stated a few posts up: My GFX-card jumps into high-load mode (full clock speed) and goes up to around 60°C in temp once I am on the login-screen (idle on the Win-desktop it's around 32-34°C). The lobby/mechlab haven't even loaded at this point. Or maybe they have and that's why the GPU is switching to max.

But even it that's the case: Setting the 60 FPS-limit hasn't changed this behaviour on my system. Once I press "PLAY" on the very first MWONL screen, the load on my GPU and the temperature go up.

Just to put things into perspective: In the actual game, I'm getting 72°C maximum (according to EVGA Precision's monitoring tools and CPUID HW Monitor). So running the actual game seems to be only slightly more taxing than running a simple login-/lobby-screen.



S.

Edited by 1Sascha, 14 August 2013 - 12:31 AM.


#64 unreal icarus

    Member

  • PipPip
  • 24 posts

Posted 14 August 2013 - 04:36 AM

View PostTostitoBandito, on 13 August 2013 - 11:34 AM, said:

You aren't understanding, it's not the game. 100% on your GPU is 100% on your GPU. It makes no difference what application is doing it or how many frames it is rendering per second. 100% on your particular GPU and laptop is clearly very hot, which is bad. You should repaste that if you haven't. The heatsink might not have good contact from the factory. If the heatsink is even a little bit off level and not making solid contact with the processor die, temps can be astronomical. They also use garbage thermal paste, and you can do much better with quality compounds.

I run a 6990M in my laptop, which is very similar to your card. When I first got my machine a couple years ago peak GPU temps in FurMark were in the mid 80's. I repasted the GPU (and CPU), replaced and added some thermal pads on the GPU, and since then my peak temp is around 73C. The cooling in your particular laptop model can make a difference as well. Mine is very robust (Alienware M17x R3) with quite a large fan hooked up to multiple heat pipes off the heatsink, and same for the CPU on the other side.

Again, you can keep blaming the game if you want but it is not the problem. Playing demanding games on laptops at safe temperatures often requires additional custom work. You can't just take a laptop out of the box and expect it to be ready for this. That is why so many people have heat-related failures and issues with laptops that they game on. They don't take those steps to lower temps and improve cooling.

http://forum.noteboo...ide-w-pics.html
That is a good guide for repasting. It was made for my laptop line, but I'm guessing yours is probably not too different.


you don't get what i mean i think. ofc heavy gpu and cpu workload produces heat. What i mean is that mwo doesnt need that much resources. i think its just a bad implementation or something like the fps thing in mechlab. I will not open my new laptop case since this would leed to loosing my guarantee.

+1 on 1sascha's post.

#65 TostitoBandito

    Member

  • PipPipPipPipPip
  • 172 posts
  • LocationWashington, USA

Posted 14 August 2013 - 06:26 AM

Well at the very least you should be able to force vsync through Catalyst Control Center. Just choose to force it to on all the time, or unless application says otherwise. This is what I do and my frame rate is capped at 60 in the mechlab from what I recall.

As for laptop warranties, opening it up and swapping components or even repasting will not usually void your warranty. It definitely doesn't in Dell/Alienware's case. Their reps openly encourage you to do so on online forums. If you are careless and happen to damage something in the course of doing this, then yes you can void it, but if you are careful you don't have anything to worry about. I wouldn't be comfortable gaming for extended periods on any laptop that I can't open up and clean out regularly, at the very least.

#66 FluffyBunbunKittens

    Rookie

  • 3 posts

Posted 15 August 2013 - 01:20 AM

View Post1Sascha, on 13 August 2013 - 09:15 AM, said:

I can only speak for my system, but even with the max FPS thing in place, I'm still getting significantly higher temps than at system idle.


Yeah... After two days of playing with the MaxFPS setting in place, things are better than they were, but MWO is still managing to tax my machine more than any other, way more visually-intensive game.

Of course, it does say Nvidia in the opening logos and I'm sporting an AMD, so perhaps they just REALLY mean that... :P

Edited by FluffyBunbunKittens, 15 August 2013 - 01:20 AM.


#67 1Sascha

    Member

  • PipPipPipPipPipPip
  • 401 posts
  • LocationMunich, Germany

Posted 15 August 2013 - 11:53 AM

Quote

Of course, it does say Nvidia in the opening logos and I'm sporting an AMD, so perhaps they just REALLY mean that... ;)


LOL.. tell that to my NVIDIA-card... :)

Minor update: I finally received all the components and put together a new system. It's a far cry from the old Core2Quad-rig of mine (i5 4670K, 2x4 GB DDR3, GTX660, Win 8 64 with an SSD boot-drive thrown in for good measure). It also resides in a brand new, much more modern and way better ventilated case *and* is equipped with an H60 CPU-watercooler.

However: MWONL's behaviour is still very noticable. The card doensn't get nearly as hot on the login-screen/lobby as the old GTX 460, but it still kicks into maximum-GPU-clock mode and gets around 10°C warmer than at idle once I click "PLAY" on the first screen.

S.

#68 Benden

    Member

  • PipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 133 posts
  • LocationFrance

Posted 26 August 2013 - 03:16 PM

View PostTostitoBandito, on 13 August 2013 - 11:34 AM, said:

You aren't understanding, it's not the game. 100% on your GPU is 100% on your GPU. It makes no difference what application is doing it or how many frames it is rendering per second. (...)


You're clearly new in the PC gaming universe if you really think that.
Just found that in 10 sec of googling :
http://www.gosugamer...d-how-to-fix-it
https://forums.stati...verheat.139827/

I've seen this kind of overheating/overdemanding bugs in pc games for over a decade. MWO behavior in the lobby is abnormal and is the result of a bug(s) or poor code optimisation at very least.

So far solutions found are forcing vertical refresh and the "Posted ImageDesouz, on 21 July 2013 - 08:51 PM, said:

I also had problems with high temperature in the mechlab
the solution is to add 'sys_MaxFPS = 60' in the config.cfg file
located in the game folder.

"
About game engine, it's clearly over-demanding : high hardware requirements for the result. But well, that's another story.
Now stop polluting this thread with non-sense or same info.
Dam, I'm polluting too ... sigh...

Edited by Benden, 26 August 2013 - 03:19 PM.


#69 Dragoon20005

    Member

  • PipPipPipPipPipPipPip
  • 512 posts
  • Facebook: Link
  • LocationSingapore

Posted 26 August 2013 - 06:28 PM

View PostDestined, on 26 August 2013 - 05:50 PM, said:

I'm just going to throw this out here...

We (IGP and PGI) are not responsible for any alterations or modifications you make to your computer, nor are we responsible with anything outside of our game. If you are doing some sort of crazy Frankenstein experiment with your hardware (or using a 3rd party program) we can not troubleshoot it.

TL:DR
Internet responsibly.



but we wont be doing any tinkering with our system if the game keeps crashing the PC or for that matter overheating the CPU/GPU when they run perfectly fine with games like Crysis 3 which uses CryEngine 3

ever since the 12 vs 12 patch

many user are suffering stability issues and drastic framerates drop with GPU overloading and crashing the whole desktop

mind you that we are running pretty decent spec PC here


edit: you should be thanking us techies who are saving you the trouble of troubleshooting stability issues

Edited by Dragoon20005, 26 August 2013 - 06:30 PM.


#70 Alekzander Smirnoff

    Member

  • PipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 427 posts
  • LocationUS

Posted 26 August 2013 - 08:15 PM

I have an EVGA GTX 760 with the ACX cooler, the temperatures at desktop for an hour (idle) it sits at 39-40C (ambient is ~21.1C). The desktop core clock is 135mhz. The splash screen of MWO it remains at that clock rate. Once I hop into the mech lab it bumps up to about 675mhz and the temperature ends up around 55C (1920x1200, very high, low motion blur vsynch off). Ingame it bumps up to 1202mhz, temperature hits maybe 67C since the rounds don't last long enough to get a full heat load.

As it is, the mech lab is a fully rendered scene with a massive UI layered on top of it. A lot of things are happening on the screen so its reasonable to say the least that it will put some load on.

On a side note my old GTX 295 which I moved up three spaces (my case has the board mounted upside down on the left hand side of the case) would constantly get clogged by cat hair and require cleaning out every so often, it would regularly get up to 80-90C when it needed a clean out. Might want to keep a can of air handy.

#71 Dragoon20005

    Member

  • PipPipPipPipPipPipPip
  • 512 posts
  • Facebook: Link
  • LocationSingapore

Posted 26 August 2013 - 08:33 PM

many user have pointed out that when playing Crysis 3

the average temps on the GPU is only around 70-79 degrees

this game

i don't even





2 user(s) are reading this topic

0 members, 2 guests, 0 anonymous users