any way to test a working GPU for issues?

goose579

Prominent
Jan 18, 2018
14
0
510
i have been browsing for a long time now and have had no luck with a diagnosis.

iv had a rig since 2011, upgraded almost all parts at this point, mobo, gpu, HDD/SSD, ram, CPU and a few others. back in February i did a clean install of the OS that didnt help much or made the issues worse. also i have uninstalled every driver and windows update i can and reinstalled from factory or driver website including beta versions and older versions.

iv had several BSOD with multiple errors codes. occasionally during BS it will have an annoying buzzing sound. sometimes when it reboots it goes to onboard graphics (crappy res, no dual monitor just mirrors) and when i go into device manager the video card is shown as not working. i disable the device then enable and it works just fine. i tried a few BSOD reporting tools in the past and posted in a different forum for BSOD and got no solid info except for clean install, update drivers, check ram. also used a monitor to track power consumption and no irregularities.

hard to remember what they all the BSOD codes are but a common "fix" that i keep seeing for most of them is: update video drivers, clean install, check ram. i have run every check i can find on the ram and have even swapped them and they all work or dont cause more issues at least. clean OS install (painful) and redone every driver i can.

is there any way to test the GPU if its faulty? (even though it works most of the time even with FO4, Skyrim, HTV Vive with no issues) the BSODs happen at completely random times. nothing makes it worse, browsing in MS Edge/chrome, playing VR games, LoL, FO4, or just writing in MS Word. nothing i do makes it worse or better and it can happen either in 20 minutes or a week.

setup:
Win 10 build 17134.228
ASUS M5A99FX Pro R2.0 AMD 990FX
AMD FX-9370-NOT OC
R9 Fury X-NOT OC
G.SKILL TridentX Series 32GB (4 x 8GB) DDR3 2133 (PC3 17000)
1000W PSU (only thing i havent needed to r2 (replace) still works great.

massive airflow with the case, antec df-85. only ever had a minor CPU overheat problem with Mass Effect Andromeda but dropped a few CPU cores when i played it and it ran fine. never seen the GPU go over 55 degrees i think. CPU has hit 60 a few times and froze but only on MEA until i dropped cores. issue started before that though.

any help would be appreciated. its VERY aggravating to start an online game to just BSOD (though SSd boot time helps) before my upgrade of the MOBO, CPU, GPU, RAM, there was no issue.
 
Make and model of your PSU? How Old is it? It seems to me, from the way you've stylized the post, that the PSU is unable to deliver the necessary power to your entire machine and it's what causes you to sometimes see the GPU.

How do you test it? You can take it to your neighbors or your friend's house and see if the card work on their system(provided they have ample power to drive the entire system with your card populating their system).

Are you on the latest BIOS for your motherboard? Might also want to see what the temps on the VRM ehatsinks are like.
 


My old setup (mobo, cup, gpu) from before my new stuff actually had a higher pull than what iv got now. Was running dual GPUs that pulled close to 200W more. I'd have to pull some stuff out to check make and model ill do that tomorrow. But it was one of the best/newish/highest rated back in 2011.

The latest bios for the mobo is from 2015 I think and iv tried older versions as well.

Iv ran a laser thermometer on every component I can see as well as using CPUID and a few other monitoring programs. Temps never got back except for the MEA game i mentioned. Is there another way to check the temps?

Its just with most of the BSOD error codes along with the "reboot with OB graphics" a lot of things end up pointing to the GPU.
 
since its been close to 7 years with this PSU should i just get a new one? the PSU calculators iv used suggest over a 1700W PSU and i think that is a bit much.

asus mobo as mentioned
R9 Fury X with liquid cooler 1 fan NOT OC
amd FX-9370 that included a liquid cooler (2 fans) NOT OC
G.SKILL TridentX Series 32GB (4 x 8GB) DDR3 2133
1 SSD
2 HDD (could remove one maybe)
6 fans in the tower each has a built in LED
a few peripherals but only my gaming keyboard and mouse are USB powered.
no disk drives

https://smile.amazon.com/dp/B017ICWP82/?coliid=I2NKC8JGNK02VC&colid=2QT8JSA3KSHDS&psc=0&ref_=lv_ov_lig_dp_it

this EVGA 1000 GQ, 80+ Gold 1000W seems to have good ratings. or any other suggestions?
 
New update,

Not the PSU. Just swapped it out (figured it was time) with a top rated 1200w. Went 3 days with no BSOD. Just had a irql_not_less_or_equal pop up when I watched a bit of YouTube. When the comp rebooted it was to onboard graphics. Disabled GPU and then enabled it. Works fine again like always.

At this point might just scrap the GPU and get a new one, I dont think I'm within any warranty anymore its been a while. Im trying to find a point of contact for AMD but all I can find is partner information. The subsystem ID shows AMD is the manufacturer......