GPU being throttled?

Hi, My name is Colin and I'm having a problem with my GPU, it almost seems as though It is not being used. this is a reoccurring and disappearing problem, it only happens every once in a while and then it will disappear and my graphics card will work fine again. I'm going to mention this now because I definitely am NOT a computer wiz. I have a little experience with computers however not enough to solve this, so this very well might not be my GPU it could be something else I have just made an educated guess that it is. Here are my specs, ive got an EVGA GeForce GTX 970, MSI gaming motherboard Z170A M5 LGA 1151, Corsair RM series RM850 power supply, an intel core I5-6500 quad core 3.2 ghz, and my ssd is a Samsung 850 Evo M.2 500GB (edit* Vengeance LPX 18 gig ram). Ive tried Completely reinstalling windows and re downloading all my drivers, ive tried changing control panel settings, I have completely taken the computer apart and rebuilt it and cleaned it. I am EXTREMELY frustrated and confused because the computer practically teases me with a few days of good 60 FPS gaming and then it poops out on me and ill be running around 40 frames on easier games and 10-20 on harder to run games like ARK or Fallout 4. Before I reset my computer I had EVGA precision X and I would monitor my stats before it would work and after. while I was playing it seemed as though my GPU use was at about 16 to 24%, however when it wasn't in a dry spell the usage would be at 80 to 100 and I would run everything fine. I monitored this while playing BF1, the computer can run what I'm throwing at it, its almost as if it just, doesn't want to​. I DONT KNOW WHAT TO DO!!!! please help, if I haven't provided enough information say so and ill try to get it to you ASAP. Thanks

Edited by Scout3108, 12 February 2017 - 11:57 AM.

Just to add, the computer just started working well again, however i noticed an odd sort of buzzing noise coming from the gpu, I have no idea what this could be, both fans are running and it seems to be working properly. (stopped working*)

Edited by Scout3108, 12 February 2017 - 02:27 PM.

Hello Scout3108 welcome to the forum,

 

What operating system are you using? I would presume Windows 10. However in my hands (or shop) the first thing I would do is take your entire system as is and duplicate your issue with the same game/s before anything else. Once I saw it happen I would see if this happens with other games as well, and monitor the GPU within windows in idle states too (the clock rates). I would also take note of the +12v rail reading in the MSI health section, and then again by using HWMonitor with GPU-Z up as well to watch clocks as I would apply various load situations. Once the issue was fully duplicated I would then simply un-attach the GPU power connector/s, and unseat the GPU, blow it out with air, reseat it, and reseat the GPU firmly. If same behavior again I would swap out the PSU with a known good working PSU and try again, and still same I would swap out GPU (another NVidia model) that is known good working and try again. If this happens again with a different GPU and different PSU then your issue would lie with the board, or operating system errors, but mostly likely the board itself. If the situation was corrected with your original parts by swapping PSU then you likely have an iffy, or bad PSU. If it was corrected by swapping out just the GPU then you likely have a bad GPU At that point I would go ahead and uninstall  the driver you are using and use DDU in safe mode and clean it, do a reinstall of a different driver (clean install) and try again with your original PSU/GPU. If happens still... likely a bad GPU. Past that I would check to see if your board had a BIOS update, or MSI may have a BIOS update for the specific MSI GPU model you have as well. If the card is still in warranty I suggest you would use it past that point. Also remember as a test you can also run your Intel I-GPU to rule out a bad PSU/GPU as well that is one of the nicer benefits of owning an Intel processor you always have that I-GPU as a fall back while you are testing, or RMA'ing a GPU.

Hello Scout3108 welcome to the forum,
 
What operating system are you using? I would presume Windows 10. However in my hands (or shop) the first thing I would do is take your entire system as is and duplicate your issue with the same game/s before anything else. Once I saw it happen I would see if this happens with other games as well, and monitor the GPU within windows in idle states too (the clock rates). I would also take note of the +12v rail reading in the MSI health section, and then again by using HWMonitor with GPU-Z up as well to watch clocks as I would apply various load situations. Once the issue was fully duplicated I would then simply un-attach the GPU power connector/s, and unseat the GPU, blow it out with air, reseat it, and reseat the GPU firmly. If same behavior again I would swap out the PSU with a known good working PSU and try again, and still same I would swap out GPU (another NVidia model) that is known good working and try again. If this happens again with a different GPU and different PSU then your issue would lie with the board, or operating system errors, but mostly likely the board itself. If the situation was corrected with your original parts by swapping PSU then you likely have an iffy, or bad PSU. If it was corrected by swapping out just the GPU then you likely have a bad GPU At that point I would go ahead and uninstall  the driver you are using and use DDU in safe mode and clean it, do a reinstall of a different driver (clean install) and try again with your original PSU/GPU. If happens still... likely a bad GPU. Past that I would check to see if your board had a BIOS update, or MSI may have a BIOS update for the specific MSI GPU model you have as well. If the card is still in warranty I suggest you would use it past that point. Also remember as a test you can also run your Intel I-GPU to rule out a bad PSU/GPU as well that is one of the nicer benefits of owning an Intel processor you always have that I-GPU as a fall back while you are testing, or RMA'ing a GPU.

 
Hi Zone_86, thank you for taking the time to respond to me, I really appreciate the help. At the time I do not posses another graphics card, but I may be able to borrow one from a friend. I understand you seem to believe the first two things I should check are the PSU and the GPU, as for cost wise I'm hoping the PSU is the one to bite the dust because I bought my GPU second hand from a relative so a warranty is not available on that one. Do you think there is anyway possible to check if the GPU is bad using a third party program such as EVGA precision? I don't know that I can borrow a PSU either however I will see about finding someone who has one. I've done some experimenting using EVGA precision and I've taken screenshots of two different games, CSGO and rocket league, tell me what you think and if you want me to take more of different games, these both are from an example of the GPU not working. (And I am using windows 10) *edit just realized screenshot didn't capture precision X and I'm slapping myself in the forehead XD

Edited by Scout3108, 13 February 2017 - 10:58 PM.

 You're certainly welcome for the response Well, more than likely from what you have said I wold lean towards the GPU and most certainly that would be the one thing to test first since you know already that it's not been working correctly based on several behavioral variables. Test your GPU in another PC and test another GPU in your PC this will very quickly tell you if it's the GPU. Sometimes it can be a driver but based on what you have said I think it's just a wonky GPU. If you have Skyrim try that one, and also Dying Light. Mass Effect would be another good one to try.

Edited by Zone_86, 14 February 2017 - 12:32 AM.

Well yesterday I went to my buddies house and tested my GPU in his computer, which he claimed could not run even the simplest games such as G-mod. we stuck my GPU in it and it started running everything fine, 60 FPS. I was shaking my head so confused, does this mean it is our PSU or something motherboard related? ive reset my computer and reinstalled windows to get rid of viruses so I don't think it could be anything like that unless its embedded. Any new ideas? should I test my power supply in his PC now WITH the GPU at the same time? Obviously my PSU isn't dead, and it seems to be powering things relatively well, however it comes to my mind that the PSU is installed with the fan facing up towards the top of my tower directly under another fan, is it possible something could have fallen in and damaged it? or would that lead to more existential damage? Still confused thanks for the ideas. I also recently replaced the Thermal paste on my GPU when I cleaned it for the 3rd time.