Problem with my graphics card?
By
Back_Blast,
Posted on: 2010-03-20 24:25 in ALL ACCESS Gameplay Technical Issues Bugs
Quote:
Shutting down is most likely due to heat in your case, but it may not be the GPU alone. Your whole system could be running hot. You can download speedfan to see your core temperature and hard drive temps, etc and if your fans are going at a good clip. See if you have system fans that are not working or a CPU fan that's stopped. If your video card has a fan, check if it works.
|
Originally Posted by Soul Storm
I recently cleared out my computer of all the dust etc. Today it has started to shut itself down when I am in the middle of playing the game.
I downloaded GPU-z and it said when loaded it was running at 102/103 degrees celsius, which doesn't take a genius to realise something is wrong here. I've emailed the people I got it from but the thing is that it's 2 years old. Will I have to pay for another replacement? Is that definitely the source of my computer shutting itself down? HELP! |
Quote:
|
Originally Posted by eltonio
This is over the top mental slavery.
|
Does your video card have a fan on it? Is the fan turning at a normal steady speed?
If you used compressed air to get rid of the dust and you didn't do something to keep the fan from turning it is possible you damaged the fan or dried out the lubricant it had that allowed it to turn freely.
If the game spit out 20 dollar bills people would complain that they weren't sequentially numbered. If they were sequentially numbered people would complain that they weren't random enough.
Black Pebble is my new hero.
My processor is an Intel i5 750 and for months i've safely had it on a turbo profile which raises the clock in hi-load situations from stock 2.6 to 3.6 ghz. There is very good airflow through the case and so the cpu temps are 30 degrees at idle and 50 at load.
The graphics card temperature as I've said is 102/103 degrees at load and around 40 at idle. I tried speedfan but it didn't come up with a gpu sensor. The video card fan works and I cleared it out of any dust etc before I took the sensor temperature reading.
The thing I don't understand is that its worked for a good 2 years why would it just stop now? The Graphics card is the oldest thing on my system, everything else was replaced just before christmas.
Edit: The fan is turning just fine still. Still as loose as the day I got it.
Quote:
102 shouldn't be the thermal ceiling for shutdown on any modern card, the threshold is usually around 120+. So I have to wonder if it's the GPU at all. I've seen video cards operate fine at around 110 before without shutting down the system.|
Originally Posted by Soul Storm
The graphics card temperature as I've said is 102/103 degrees at load and around 40 at idle. I tried speedfan but it didn't come up with a gpu sensor. The video card fan works and I cleared it out of any dust etc before I took the sensor temperature reading.
|
What model card do you have? What happens when shut down occurs, is it like hitting the reset button, is it a hard power off, or is there an error message on the screen? Does it only happen in the game?
Quote:
| The thing I don't understand is that its worked for a good 2 years why would it just stop now? |
Things break.
Quote:
|
Originally Posted by eltonio
This is over the top mental slavery.
|
It happens when I am playing CoH (it's happened 3 times this afternoon only when playing CoH) and never any other time. I can be happily playing away and suddenly the display goes blank and the computer has turned itself off. It's not a restart or anything, it's a hard power off. You can turn it straight back on again no problem.
When this happened earlier I took a look inside the case, the CPU cooler (general area) didn't feel too hot. The GPU I could feel it was quite hot. It cooled itself down after a few minutes, but the fan works fine so I don't really get why its doing it.
There is no warning, no artifacts nothing to suggest something is going to break or go wrong and then boom suddenly the computer is off.
What video card do you have?
As far as "it doesn't do it with other games", that's not really a valid argument. CoH stresses a system, and especially the graphics card, far more than any other game I've seen. I've used it to stress test a system. I've seen reports of others using CoH as their test for how well a computer runs.
It sounds like a probelm with the video card, but without being able to swap it out into another computer or swap in another video card I can't tell for certain.
If the game spit out 20 dollar bills people would complain that they weren't sequentially numbered. If they were sequentially numbered people would complain that they weren't random enough.
Black Pebble is my new hero.
Sorry I knew there was something I forgot on that last post. I have a ATI Radeon 3850 Pro 512MB card. I run COH on 1600x1050 with everything maxed out. Have done ever since I got the card. I play other newer games like TF2, Borderlands and @*cough*apb*cough* (you didn't hear me say that!) All completely maxed out and none have ever made this problem occur.
Quote:
Yep I asked this too.|
Originally Posted by Texas Justice
What video card do you have?
|
Quote:
| As far as "it doesn't do it with other games", that's not really a valid argument. CoH stresses a system, and especially the graphics card, far more than any other game I've seen. |
Quote:
|
Originally Posted by eltonio
This is over the top mental slavery.
|
Quote:
I concurr with this statement. |
Originally Posted by Smurch
CoH is one of the least performance intensive games I play.
|
Another thing that might be relevant is that mid-week I updated my bios. Perhaps that has affected something? Should I roll-back to the previous version?
Quote:
The GPU will feel hot to the touch, certainly. That's not a good gauge for anything. 102 shouldn't be shutting the system down. |
Originally Posted by Soul Storm
It happens when I am playing CoH (it's happened 3 times this afternoon only when playing CoH) and never any other time. I can be happily playing away and suddenly the display goes blank and the computer has turned itself off. It's not a restart or anything, it's a hard power off. You can turn it straight back on again no problem.
When this happened earlier I took a look inside the case, the CPU cooler (general area) didn't feel too hot. The GPU I could feel it was quite hot. It cooled itself down after a few minutes, but the fan works fine so I don't really get why its doing it. |
I am not sure if Radeon has something like NVIDIA's Nfuse, but if so, download it and have it run a stress test against your video card. Some burn-in testing software also has this capability. See if you can get the system to fail and under what conditions.
You can also try turning off thermal shutdown in your BIOS, if that's supported. It's usually in the Advanced Power options. If you turn that off, and it still shuts down, then you've got another issue. You'll want to run Speedfan or another monitor to warn you if your system temp or GPU temp approaches the normal shutdown threshold for your model, however, and then shut down manually, so be careful when trying this and keep an eye on the temp.
Quote:
|
Originally Posted by eltonio
This is over the top mental slavery.
|
Quote:
Whats the peak temp of your GPU while playing those other games?
|
Originally Posted by Soul Storm
Sorry I knew there was something I forgot on that last post. I have a ATI Radeon 3850 Pro 512MB card. I run COH on 1600x1050 with everything maxed out. Have done ever since I got the card. I play other newer games like TF2, Borderlands and @*cough*apb*cough* (you didn't hear me say that!) All completely maxed out and none have ever made this problem occur.
|
Quote:
|
Originally Posted by eltonio
This is over the top mental slavery.
|
I've tested some games. If I run CoH in windows mode the temperature goes to 90 degrees. In Borderlands the temperature floats between 80 and 100. Tf2 was the same if a little less at more or less 75-85 most of the time.
A friend of mine seems to think reverting the Bios to the version I had a week ago will help the problem. As far as trying to force the card to crash out again I'll try it tomorrow as its quite late here now.
Thanks for the help so far guys. Is it looking to be a specifically graphics card thing? The card is 2 years old only but I don't think the place I got it from would replace it now would they?
Quote:
I didn't say it was performance intensive, I said it will stress a system, especially the graphics. It has more to do with the Cryptic engine's coding than anything else from what I can tell, but it is still a good stress test for a computer.|
Originally Posted by Smurch
Yep I asked this too.
You must not play very many newer games. CoH is one of the least performance intensive games I play. |
Also, TF2, Borderlands and AGP all seem to be DirectX games. City of Heroes is OpenGL. It's like comparing apples to oranges.
If the game spit out 20 dollar bills people would complain that they weren't sequentially numbered. If they were sequentially numbered people would complain that they weren't random enough.
Black Pebble is my new hero.
*bump*
Is the only way I'm gonna get this sorted to buy a new card!?
A new card isn't necessarily the solution. If you want to try and eliminate the card as the issue, I'd suggest seeing if someone you know has a comparable card you can borrow for testing and try using that in CoH and see what happens. That way you don't shell out $$ for a card only to discover it's something else causing the problem.
And you could try rolling back the BIOS as you've considered though I'm not sure that will do anything. I'd try a card swap first though.
Also, I'd tend to guess you can't take the card back to where you got it. But if it is still under warranty from the manufacturer and you can show that the card is faulty, then you may be able to get the maker to do something.
It is known that there are an infinite number of worlds, simply because there is an infinite amount of space for them to be in. However, not every one of them is inhabited. Therefore, there must be a finite number of inhabited worlds. Any finite number divided by infinity is as near to nothing as makes no odds, so the average population of all the planets in the Universe can be said to be zero. From this it follows that the population of the whole Universe is also zero, and that any people you may meet from time to time are merely the products of a deranged imagination.
Quote:
Not sure if you mentioned this already but what is the temp of the CPU when you run COH? You said you removed the dust from the computer, was there a lot? If so, did you clean the heat sink of the cpu also?|
Originally Posted by Soul Storm
I recently cleared out my computer of all the dust etc. Today it has started to shut itself down when I am in the middle of playing the game.
I downloaded GPU-z and it said when loaded it was running at 102/103 degrees celsius, which doesn't take a genius to realise something is wrong here. I've emailed the people I got it from but the thing is that it's 2 years old. Will I have to pay for another replacement? Is that definitely the source of my computer shutting itself down? HELP! |
As other have said, the graphic card running at 102 shouldn't trigger the computer to reboot. On the other hand, the CPU running too high (high temp limit set too low) would do just that. You changed the bios recently, did you make sure that all the settings for the temp monitor were set correctly?
When I was running CoH fullscreen it was 102/103 degrees. Ever since I've been running CoH in window mode and it still heats up to 85-95 degrees but it's not forced a hard reset. I've heard it might be something to do with the Thermal Paste on the card drying up but I don't really know much about that. I tried to get the shop where I got it from to replace it but they said it's out of warranty. I also tried to contact ATI but I've not heard anything back from them since.
I'd like to try out a different graphics card but I've not got another and no one I know has one comparable. When I did the dusting of the computer the other day there was little to no dust surprisingly. The only place there really was any was in the graphics card cooler and like I said I've cleaned that out now.
Thanks for the help so far guys, just want to get this sorted.
If your CPU and video card fans are spinning great, but your power supply most likely has one as well, If it dies you'll have a problem very nearly the same as with the other two.
@MARTy McFly
I don't know if the following is helpful but here are my full specs and a small update:
Graphics Card: ATI Radeon Hd3850 Pro 512 MB running 10.2 drivers.
Mobo: Asus p7p55d Pro
Memory: 4GB G. Skill Ripjaw Dual Channel 1600Mhz Kit DDR3 PC3=12800C9
PSU: Hec-550td-pte 550W
Case: From Maplin, http://maplin.co.uk/Module.aspx?ModuleNo=97188
Processor: Core i5 750 originally overclocked using the bios turbo profile to 3.7ghz from the stock speeds, and reliably so for a good few months, with cool idle temps of 32 degrees Celsius and loads of 50 degrees. But since this graphics problem I returned it to stock speed just in case. The fact that I have been able to run it like that for so long I would think suggest there’s not a problem with airflow throughout the case.
I also ran the Furmark test, and the card started off at 72 degrees Celsius and went up to 119 running with Post FX and Displacement Maps with 8x MSAA at 1680x1-5- resolution
My recent experience is nearly identical to Soul Storm's. Been running CoH on my system for 2 years with nary a glitch, but the last two weeks it's been 'Locking up' during missions. Screen freezes and then nothing works, requiring a power off/on.
I cleaned out my PCs innards last night and it was pretty dusty, but I was thorough. Even so, soon as I logged in and attempted a mission...Lock Up!
And like Soul Storm's, nothing else I run causes any problems...
Could be my card's going bad. I'll try some of the monitor things you guys are suggesting, and hope I don't have to buy a new video card. Swapping one out to test isn't an option.
I just ran the IntelBurn Test on the standard stock speed everything was stable. Then I changed the profile in the bios to Turbo, which overclocks the CPU to 3.71Ghz and the IntelBurn Test said the system became unstable (I'm running the standard stock cooler, If I wanted to run with this setting normally I'd have to upgrade probably the cooler and case?) which is what the system was on when the above problem started.
I have since changed the profile to Better Performance which runs the CPU @ 2.97Ghz and everything is proved stable CPU wise. Would this the be source of my problem or is the graphics card still at fault? The GPU Stress test I posted above does suggest a high temperature of 119 degrees.
Well an unstable CPU can cause issues. So I'd imagine it's at least possible that is the root of your problem. Run it at the lower or even stock speed for a bit and see what happens. As for why it is unstable, the stock cooler being insufficient could be it but also it's possible the BIOS profile used is not quite right for your chip. Overclocking is not an exact science. Tweaking the profile might render it stable at that speed, or it might fry the chip altogether. OC at your own risk.
It is known that there are an infinite number of worlds, simply because there is an infinite amount of space for them to be in. However, not every one of them is inhabited. Therefore, there must be a finite number of inhabited worlds. Any finite number divided by infinity is as near to nothing as makes no odds, so the average population of all the planets in the Universe can be said to be zero. From this it follows that the population of the whole Universe is also zero, and that any people you may meet from time to time are merely the products of a deranged imagination.
Quote:
Yes, run some monitors and check your temps, fan speeds and the like. I'd also make sure everything is properly seated and such. It might be nothing more than a loose connection you created while poking around in it to do your cleaning.
|
Originally Posted by Socorro
My recent experience is nearly identical to Soul Storm's. Been running CoH on my system for 2 years with nary a glitch, but the last two weeks it's been 'Locking up' during missions. Screen freezes and then nothing works, requiring a power off/on.
I cleaned out my PCs innards last night and it was pretty dusty, but I was thorough. Even so, soon as I logged in and attempted a mission...Lock Up! And like Soul Storm's, nothing else I run causes any problems... Could be my card's going bad. I'll try some of the monitor things you guys are suggesting, and hope I don't have to buy a new video card. Swapping one out to test isn't an option. |
It is known that there are an infinite number of worlds, simply because there is an infinite amount of space for them to be in. However, not every one of them is inhabited. Therefore, there must be a finite number of inhabited worlds. Any finite number divided by infinity is as near to nothing as makes no odds, so the average population of all the planets in the Universe can be said to be zero. From this it follows that the population of the whole Universe is also zero, and that any people you may meet from time to time are merely the products of a deranged imagination.
I recently cleared out my computer of all the dust etc. Today it has started to shut itself down when I am in the middle of playing the game.
I downloaded GPU-z and it said when loaded it was running at 102/103 degrees celsius, which doesn't take a genius to realise something is wrong here. I've emailed the people I got it from but the thing is that it's 2 years old. Will I have to pay for another replacement?
Is that definitely the source of my computer shutting itself down? HELP!