Jump to content

Welcome to Geeks to Go - Register now for FREE

Need help with your computer or device? Want to learn new tech skills? You're in the right place!
Geeks to Go is a friendly community of tech experts who can solve any problem you have. Just create a free account and post your question. Our volunteers will reply quickly and guide you through the steps. Don't let tech troubles stop you. Join Geeks to Go now and get the support you need!

How it Works Create Account
Photo

CPU and GPU running hot!


  • Please log in to reply

#1
rhymin

rhymin

    Member

  • Member
  • PipPipPip
  • 210 posts
I downloaded CPUID HWMonitor to check my PC temps (I'm not sure why I didn't do this a year ago, but the loud fans have been driving me crazy lately so I wanted to get more involved). Anyways, I found that my CPU temp (all 4 cores) are running between 56C-63C, and my 2 GPUs are running at 79C-88C. I am not running any games or any big application right now, and this is what they are currently reading. Actually, in the HWMonitor, I only see 1 video card showing up, so I'm not sure if this is the temperature of 1 or both.

My system specs are as follows:

GPUs - Radeon HD 5770 (I have 2 of these running crossfire)
CPU - AMD Phenom II x4 965 (Not overclocked and all 4 cores run at the same temp)
MOBO - ASUS M4A79XTD EVO
Case - AZZA Solano 1000 Black FULL ATX
PSU - Corsair TX Series CMPSU-750TX 750W

I put this machine together in 2010 and used the factory thermal paste, but I am going to apply my own tomorrow when I get time. I'm hoping this will drop the temp by at least 10C. As for the GPU, I have no idea how to cool that down, any suggestions?
  • 0

Advertisements


#2
rhymin

rhymin

    Member

  • Topic Starter
  • Member
  • PipPipPip
  • 210 posts
Actually, I just downloaded GPU-Z, and it shows 1 video card running at around 90C, and the other running at 50C. The fan speed is also at 3300 RPM for the first one, and 1400 for the second. Also, it doesn't show the memory usage stats for the 2nd card, but shows 276 dedicated memory usage for the first one. I'm not sure why 1 card is running so much hotter than the other. Does this have anything to do with crossfire not being enabled correctly, or 1 card possibly being faulty?
  • 0

#3
iammykyl

iammykyl

    Tech Staff

  • Technician
  • 7,659 posts
Gday.
Dust and dirt are probable the main factor in normal overheating, new TIM may help and a good clean is in order especially the video cards. Take anti static measures before working on the computer and when handing parts.
The card that is connected to the display will run hotter as it is doing more work and should be in the primary PCI-E slot. Mark the top card with a pen and then take them out, remove the shroud, use cotton buds to loosen stubborn dirt, keep a finger on a fan blade and use compressed air to blow them out, also blow out the slots. replace the cards and test.
Here is one article, I would search and read more. > http://smallbusiness...card-69623.html
  • 0

#4
rhymin

rhymin

    Member

  • Topic Starter
  • Member
  • PipPipPip
  • 210 posts
That article says to take the video card completely apart and clean the heatsink and apply thermal paste. I've never taken apart a video card and am afraid that I could damage it in doing so. Why would one video card be running so hot and the other one not at all?

Also, thank you for the quick reply!

Edited by rhymin, 29 December 2013 - 11:48 AM.

  • 0

#5
painstream

painstream

    New Member

  • Member
  • Pip
  • 4 posts
Its crazy how much a little bit of dust in the fans can cause overheating issues. I would get a can of compressed air and blow the dust out of it. The primary video card typically will have a higher load than the secondary and generate more heat (swap the cards around and you should be able to replicate this behaviour). By chance are you using any video card software that is automatically over-clocking the cards?. Also I personally never ever use stock thermal paste on anything as I find the thermal paste most vendors use is garbage.. I usually will apply arctic silver 5 thermal paste and have seen temps drop as much a 8 degrees C from stock paste to the new paste.. (applying paste too thick will cause heat issues as well). Also if your still stuck see what happens when you just have 1 video card in your system. I usually find SLI rigs run hotter than stand alone GPU's.
  • 0

#6
iammykyl

iammykyl

    Tech Staff

  • Technician
  • 7,659 posts
Gday.
I gave you the link to that article to start you off. You do not have to go as far as complete disassemble, you might try just using some compressed air to blow out the dust but i don't know how successful that would be.
Watch some videos to see how hard it might be, > https://www.google.c...HD 5770&tbm=vid

If you are running games in a window, try running in full screen.
Try swapping the cardsaroundnd to see if it makes a difference.

TIM. More care is to be taken using a paste containing metals as excess amounts can short out circuits on the MB. For less experienced people, use a non conduct one like > http://www.newegg.co...N82E16835186038
  • 0

#7
rhymin

rhymin

    Member

  • Topic Starter
  • Member
  • PipPipPip
  • 210 posts
I used compressed air to blow out a bunch of dust in both video cards. I also swapped them around as you both suggested. I am now getting around 55C and 40C in both cards, which is a HUGE dropoff from the 90-100 my first card was generating before. If one of them acts up again, I will disassemble the cards and clean the heatsinks and fans out better, as it doesn't appear to be too difficult in opening this particular card.

Question: What method of applying thermal paste to CPUs and GPUs do you guys recommend? I know to only apply a "pea-sized amount" so the layer isn't too thick, but I've seen different methods. The two most popular I have seen is to just put a dab in the middle and use the force of the heatsink to spread the paste. Another method that I have used it to spread it with a little piece of paper (which I've heard can create air bubbles). I am always afraid if I just use the heatsink force method, that it won't spread evenly or cover the entire surface. Any suggestions?

My CPU is still running at around 60-65 C when idling, which is too hot...and why I ask.

Edited by rhymin, 29 December 2013 - 09:29 PM.

  • 0

#8
iammykyl

iammykyl

    Tech Staff

  • Technician
  • 7,659 posts
That's good news for the video card.
Do not use the spread method, takes a lot of practice to get it right and you can still get poor coverage and air bubbles.
First, use a non metallic TIM then there is no chance of shorting out parts.
Second, Please have a read of this article, then reply, > http://www.pugetsyst...Techniques-170/
  • 0

#9
rhymin

rhymin

    Member

  • Topic Starter
  • Member
  • PipPipPip
  • 210 posts
I have some bad news about my CPU. It was absolutely CAKED to my heatsink. It took me 30 minutes to get off, and I guess I damaged my CPU in the process. My computer turns on and the fans spin, but I get no beeps or cursor to the monitor. It says "check cable source". I'm on my old PC now. I really wish I would have just left it alone. I even bent a couple pins in the process, but was able to straighten those out and fit it back into the socket. Still nothing. Normally I would be extremely pissed off, but I'm trying a new live, learn, and relax approach in life, lol. It's not the end of the world, but a costly mistake on my part.

Does this sound like a CPU problem? If so, what type of CPU do you recommend buying? I'm not sure they even make AMD Phenom II x4 965 anymore?
  • 0

#10
iammykyl

iammykyl

    Tech Staff

  • Technician
  • 7,659 posts
Now that is not good news but may not be a disaster.
Check that all the cables are still properly seated, especially the 12v one that runs to the top left of the MB.
Is the CPU fastened down correctly, give it a gentle wiggle, should be firm and not able to rock from corner to corner.
Remove the RAM and then reseat it ensuring the locking tabs at each end of the sticks is fully closed.
Is the video cable on the correct sockets and fully secured
Test.
  • 0

Advertisements


#11
rhymin

rhymin

    Member

  • Topic Starter
  • Member
  • PipPipPip
  • 210 posts
I re-seated every component and cable multiple times and even went with just 1 vid card and 1 stick of ram (the minimum basics) to see if it would work. I still can't get input to my monitor. The CPU doesn't wiggle at all and is firm. Every pin straightened out perfectly when I latched it in the first time. I know the monitor works because I am on it now with my old PC. I'm assuming it is my CPU, and I might buy a used Phenom II x4 965 on eBay. I see them going around $125-$150, and that is right in what I am willing to spend.

If you agree that this is a CPU problem, I am going to go ahead and buy one ASAP. Thank you for helping!
  • 0

#12
rhymin

rhymin

    Member

  • Topic Starter
  • Member
  • PipPipPip
  • 210 posts
I'm actually thinking about the AMD Phenom II x2 which is unlockable to 4 core. The seller in this ebay auction guarantees that all 4 cores are operational. Wouldn't this CPU perform the exact same as the Phenom II x4 965? I do not care about OC'ing, so that is not a factor in this decision. The reason I am wondering is because the B59 is about $100 cheaper than the 965.

Here is the auction and he will give a full refund if it doesn't work out: http://www.ebay.com/...=item1e7cb2ca11
  • 0

#13
iammykyl

iammykyl

    Tech Staff

  • Technician
  • 7,659 posts
Gday.

If you agree that this is a CPU problem, I am going to go ahead and buy one ASAP.

Can't say 100%, but is most likely. It could be damage to the MB and the only way to know for sure is if you install any known working CPU that is compatible. Bit of a catch 22.


Re the B59. Performs exactly the same in single threaded applications, performs the same with all 4 cores unlocked but, may not run stable because some 965 were binned with a defective core, some were binned due to low 2 core stock.

Just checked and your MB supports Core Unlocker.

Some checks before calling it a loss.
With the PSU off, perform a BIOS reset by using the clear CMOS jumper method, then test.
Remove the CPU, check it and the socket in case a pin has actually broken off.
  • 1

#14
rhymin

rhymin

    Member

  • Topic Starter
  • Member
  • PipPipPip
  • 210 posts
Well I didn't think I could test my mobo or CPU, but I was able to and now know that my CPU is bad...I'm just hoping my mobo isn't as well (which would be extremely rare, correct?) Also, I performed the BIOS reset and checked the pins on the Phenom CPU, and they were all there.

I wasn't able to test a CPU in my ASUS M4A79XTD EVO motherboard, because the Athlon wouldn't fit. But I was able to test my AMD Phenom II 965 on my M2N-SLI DELUXE mobo. I just heard a loud high pitched screeching sound. No beeps, and no cursor or connection to the monitor. Then I put my AMD Athlon 64 X2 6000+ back into the M2N-SLI mobo, and it booted right up.

I have a couple questions:

1) Does that high pitch screeching sound mean anything? I know that beep sequences do, but it wasn't really a beep.

2) Would you recommend me purchasing that used B59 for $60, or spending a little extra and buying a used Phenom II 965? There is actually an auction ending in less than 30 minutes for a Phenom II, and I might have to bid on that one.
  • 0

#15
rhymin

rhymin

    Member

  • Topic Starter
  • Member
  • PipPipPip
  • 210 posts
Nevermind that previous auction, but I'm guessing I'll have to pay about $140-$150 for a used 965. Would you do that, or pay about $60 for a used B59? Again, I really don't care about OC'ing, but I do want all 4 cores to run stable.
  • 0






Similar Topics

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

As Featured On:

Microsoft Yahoo BBC MSN PC Magazine Washington Post HP