Sign in with
Sign up | Sign in
Your question
Closed

Acceptable Temp Range for CPU/GPU

Last response: in Graphics & Displays
Share
August 28, 2012 9:03:32 PM

Could someone please give me a general idea what the acceptable temp range should be for my CPU and GPUs? I realize this is a pretty basic question, and I've found decent answers on the web, but unfortunately each answer fails to either
  • Explain the method to derive the temp. I know there are many different tools out there.
    or
  • It is for a different GPU/CPU.

    I am definitely a noob, I'm having some problems and I've been banging my head against the wall for several weeks. I think my current GPUs are bad and I'm wondering if they've gotten too hot in the past. I'm about to buy replacements, so I'd really like to have a rock solid range and measuring tool in place to monitor the temps of the new cards. This is the first system I've ever built and I'm wondering if the case doesn't have enough air flow to support the cards. I've always felt like the system was a little unstable, but the only game I've tried to play is CIV 5 and apparently there are tons of people having problems with it, so I never knew if it was my rig or the game. Now the problems are getting worse and occurring outside of the game, so I'm going to try to start ruling out the different hardware components, but I don't want to fry the new ones too!

    MB: Asus P6T6 WS Revolution
    Current Graphics Cards: 2 BFG Nvidia GTX 295
    New Graphics Cards: 2 Asus Radeon HD 7870
    CPU: i7

    Thanks in advance for any input you can provide.
    a b U Graphics card
    a b à CPUs
    August 28, 2012 9:15:26 PM

    First cool down! It's very unlikely that one of your components overheated in the past, because they should shut down before that (unless you have switched that off).
    For measuring the temps I recommend the tool that came with the component (MB, GPU) The manufacturer always knows best where the sensors are, what the offset is and what the acceptable ranges are. The default setting of the tool will tell you that.
    Score
    0
    August 28, 2012 9:19:50 PM

    For any correctly created modern CPU/GPU the maximum safe operating temperature is somewhere around 90C. If you reach these temperatures they should not be for extended periods of time. The maximum extended operating temperature should be at 80c or lower.

    The maximum temperature before silicone / solder damage starts is right around 100c. If your CPU or GPU ever hit 100c, then it is possible they are damaged, also if they ever ran at 95C+ for extended periods of time, then it is likely they are damaged.

    Download and run OCCT ( http://www.ocbase.com/ ) select the power supply test on your system, this will push your HDD, ram, CPU and GPU at the same time (stressing the power supply.)

    There are frequency, memory, and temperature monitors to show you whats happening with your system. Run the test for 30 minutes and watch your system the whole time. If your video starts giving you artifacts/garbage, then its probably running too hot or been damaged, if you get a BSOD, then your CPU/motherboard may be damaged. If your system locks up or restarts then its probably your power supply.

    Good luck. :) 

    Score
    0
    Related resources
    August 28, 2012 9:31:10 PM

    Thanks for the info guys. Just as an FYI, I have already downloaded OCCT. However, I didn't use the Power Supply test, but I'll be sure to try your suggestion. I can get it to BSOD after a few minutes using the GPU Test. The BSOD error message pointed to the Nvidia card so that was my first suspect. Anyways, thank you for the information.
    Score
    0
    August 28, 2012 9:31:28 PM

    Best answer selected by Johnny6.
    Score
    0
    August 28, 2012 9:51:01 PM

    Thanks for the best pick Johnny. If you are seeing a CCC*.* or NV*.* type BSOD, then try using msi afterburner to decrease your video card clocks back to stock settings. If they are already at stock, try dropping the memory clocks by 10% then test again, then chip clocks by 10% then test again.

    If dropping clocks below stock resolve the issue, your video card manufacturer should replace the cards, as long as the heat sink has not been tampered with.
    Score
    0
    August 28, 2012 10:04:15 PM

    Johnny6 said:
    Could someone please give me a general idea what the acceptable temp range should be for my CPU and GPUs? I realize this is a pretty basic question, and I've found decent answers on the web, but unfortunately each answer fails to either
  • Explain the method to derive the temp. I know there are many different tools out there.
    or
  • It is for a different GPU/CPU.

    I am definitely a noob, I'm having some problems and I've been banging my head against the wall for several weeks. I think my current GPUs are bad and I'm wondering if they've gotten too hot in the past. I'm about to buy replacements, so I'd really like to have a rock solid range and measuring tool in place to monitor the temps of the new cards. This is the first system I've ever built and I'm wondering if the case doesn't have enough air flow to support the cards. I've always felt like the system was a little unstable, but the only game I've tried to play is CIV 5 and apparently there are tons of people having problems with it, so I never knew if it was my rig or the game. Now the problems are getting worse and occurring outside of the game, so I'm going to try to start ruling out the different hardware components, but I don't want to fry the new ones too!

    MB: Asus P6T6 WS Revolution
    Current Graphics Cards: 2 BFG Nvidia GTX 295
    New Graphics Cards: 2 Asus Radeon HD 7870
    CPU: i7

    Thanks in advance for any input you can provide.



  • Johnny6

    You did NOT say what you are using for a PSU ie Power Supply, and you also said GPU's as in plural. The GTX295's are dual GPU units and you are using 2 cards or 4 GPU's that need a sizeable power source. If you answer the PSU question, the solution might be as simple as a better power source. Advise Jellwood
    Score
    0
    August 28, 2012 10:13:14 PM

    Also there is this to consider--Bad News For BFG Video Card Owners
    Aug 16, 2010 at 3:25 PM - OhmWrecker - 13 Comments
    News broke out this past May that video card manufacturer BFG Tech was leaving the graphics market, leaving many surprised at the lack of any warning. The only reason given for the exit was that the graphics card market was no longer profitable for the company. The news came with a promise however that warranty support would continue on, even for lifetime warranty products.

    Just three months later it looks like BFG Tech is whistling a different tune. HardwareCanucks reports on a letter received by one of their readers which very clearly states that video card RMA support will no longer be honored. This is despite the fact that the company continues to operate in other markets, and has a large number of customers that purchased their graphics cards with the lifetime warranty promise in mind.

    The rejection letter reads:

    BFG Technologies Inc. is in the process of winding down and liquidating its business. Unfortunately, our major supplier would not support our business. As a result, we are returning your graphics card without being able to repair it. We apologize for the inconvenience.

    Major bummer for BFG card owners. It just goes to show that the true value of a lifetime warranty is something that one should always question.

    Hope this helps. Jellwood
    Score
    0
    August 28, 2012 10:42:56 PM

    Thats for the disheartening replies Jellwood. Fortunatly Johnny stated that his current video cards are 7970's, and his previous were GTX 295's. Since BFG was never really with the amd camp, this shouldn't impact Johnny, but that still sucks :( 
    Score
    0
    August 29, 2012 12:09:42 AM

    I did not mean to be disheartening, rather solve a problem. The PSU question still is unanswered and needs to be addressed. I do BOINC and SETI and the choice of GPU is a big issue in the parallel/distribative computing world. I just purchased a New in the Box EVGA GTX295 and a Gigabyte GTX670 and a pair of Zotac GT640's to go into "Bulldozer" Powered 990FX Motherboards from ASUS, Gigabyte, and MSI. Built 40 some Mobo's in the last few years, All with AMD cpu's and nVidia gpu's, all to process SETI work units. Temps vs Watts is a big issue for me and can understand Johnny's questions. I HAVE owned BFG cards in the past and they were good. Be more than happy to buy the 295's if they are for sale. The new one cost me $100.00 Mich salestax included. I am not an Expert at anything but I do read a lot and forget little. The type of case should also be investigated, for airflow and heat buildup. Hope this helps. John Burtscher or jellwood@tir.com
    Score
    0
    August 29, 2012 9:39:36 PM

    No no it wasn't the way you replied that was disheartening so much as the content. I was sad to find out BFG is no longer honoring their warranties :( 

    Additional input is always welcome, I know I am not always right, and the more options the users have the better.
    Score
    0
    a c 271 U Graphics card
    a b à CPUs
    August 29, 2012 10:31:19 PM

    This topic has been closed by Mousemonkey
    Score
    0
    !