So my computer hasn't really had any issues. I've been able to watch video, plays games such as starcraft 2, diablo 3, skyrim, team fortress, left for dead 2, and borderlands 2 on high settings with no problems.
However as of about a week ago I've had it shut off on me twice while playing borderlands 2. Since nothing else seemed wrong I assumed it was overheating and shut off to protect it from damage.
Here are my specs
Processor: Intel® Core™ i7 950 Processor (4x 3.06GHz/8MB L3 Cache)
Memory: 6 GB [2 GB X3] DDR3-1600
Video Card: 2x - ATI Radeon HD 5770 - 1GB - CrossFire Mode (Dual Cards)
Motherboard [SLI] Gigabyte GA-X58A-UD3R w/ 4x PCI-E 2.0 x16
Using Core Temp my cpu seems to get to 66C max.
I used GPU-Z to take a look at my temperatures and 1st gpu is normally around 70C and can get up to 105C(so far as I stop playing and let it cool down). It's load usually doesn't go higher than 30-40%.
The other seems like it is hardly used. For the most part the temp stays at around 50C and its load is usually at 0-1% but can sometimes spike to 27% for a few seconds, though this is rare.
Crossfire is enabled in CCC, and I turned on the graphic when its working and that does show up when playing borderlands 2.
I tried taking out both cards, unscrewing and opening them up, using compressed air to get any dust out (though there really wasn't any), removed the old thermal paste and applied some new paste. This seems to have helped a little but it still runs hot. Now it stays in the 67C-95C range. Tried updating the drivers, but they are already updated. Tried updating CCC but its already updated.
So why does only one card get so hot? Why is only one card used most of the time? Why was I able to play borderlands 2 for over 30 hours over the course of a few weeks and have no issues and suddenly there is overheating/shut downs? What else can I try?
Yeah the top card is the one being used and the one getting hot. I believe there are 4 slots and it is in slot 1, while the other card is in slot 3. Due to the size though they are right next to each other. The fan at the bottom of card1 is right above card2. They are both positioned towards the bottom back of the case.
Also forgot to mention that both fans are working. I can physically see them moving. There are also 3 other fans running. One in the back of the tower just above card1 blowing outward. One at the top of the case blowing inward. One at the front bottom blowing inward. Removing the side of the casing doesn't help with the temp.
The idle temp difference between the 2 cards is about 20-25C and I never had an issue with it shutting down in the past using this configuration so I don't understand why it would start all of a sudden last week if its just a spacing issue.
Nonetheless I'll shut down and give it a try and report back when I can
You could try switching the cards (moving each one to the other's location) to see if the problem follows the card or stays with whatever card is in the top slot.
If it follows the card, (meaning the bottom card gets hot after you switch locations) then there is a problem with the card. If not, it's probably a configuration problem that had not gotten bad enough to shut down until now.
I suspect your top card has been running hot for a long time and just recently crossed the heat threshold where it shuts down.
Can't move the 2nd card to the bottom as there is not enough room.
It is possible it was running hot the whole time. I didn't start monitoring until there was a problem. It seems odd that I could play a 2 hour session with no issues and then have it shutdown another time in <15 minutes. It also seems odd that changing all the various graphics and options settings from high quality to low doesn't help. I would think the card wouldn't have to work as hard and therefor shouldn't get so hot.
I'll try switching the cards as you suggest.
Out of curiosity could I switch which one is the main one being used by either plugging the monitor into the 2nd card or by switching primary card through windows itself?
The theory you want to test is: does the location of the cards cause the top card to get hot, or is the top card getting hot because it some other issue?
The most direct way to test that theory, is to make the two cards switch slots.
Use the same two slots you are using now, just take the "cool" card and move it into the "hot" card's top slot....and move the "hot" card into the "cool" card's slot.
If the card that's getting hot now, (the top card) continues to overheat after you move it to the lower position ....(not the very bottom one that you have never used, but the lower slot you are using now)... The card has a problem.
Just plugging your monitor into the other card without moving them won't change the heat/airflow situation for either card. The bottom card will still be getting nice cool air and the top card will still be smothered by the bottom card.
I switched the cards and now there is no overheating at all.
Card1(Formerly Card 2) averages 70 and gets 76C at max and card2(formerly card1) averages 60 and gets up to 70 max.
I'm guessing the only reason card2(formerly card1) isn't getting hotter is because its on the bottom now and it is hardly ever under load. So it was most likely the problem. Thanks for the help.
While playing I noticed both cards were intermittently going to ~30% load, which I would assume is normal with crossfire, but most of the time card2 is under 0 load even when card1 is in the 30-35% range.
This even occurs during graphic intensive times like when I'm running through an area and fighting even more enemies at once. Yet card2 is under 0% load the entire time.
In GPU-Z card1 has Memory Usage (Dedicated) and Memory Usage (Dynamic). But card2 doesn't. It that normal?