Jump to content

*HELP* GPU troubleshooting / problems

i5-8600
Sapphire Nitro+ RX 580 8GB (1411 Core clock / 2000 Memory clock)(OC'd to 1455 and 2100)(Tested OC'd and Stock)
Gigabyte B360M DS3H
8BG of 2400Mhz RAM OC'd to 2666 Mhz


Ok guys, so long story short - My computer crashed right in the middle of a PUBG game several days ago and I had to do a hard reset. Upon reboot, my SSD which the OS was installed on had failed. You know the story, had to re-install windows via USB on my HDD because the SSD was not able to initialize using windows Disk manager. Anyway, BEFORE the "great crash of 2018" my GPU was working fine. I normally played in VSYNC at 75 frames in FPS games and let the frames run wild in Rocket League at 250 at all max settings. No issues. Card normally ran at 68-70*C, at 1455Mhz Core Clock / 2100 Mhz memory, using anywhere from 2GB to 5GB of VRAM, and on the CPU side using anywhere from 10%-12% CPU and very little RAM. AFTER the crash, it now jumps around below 1455Mhz, uses 0.7GB of VRAM on Rocket League with 4-5GB of RAM and 30-40% CPU usage. On PUBG it now uses 4-5 Gb of VRAM and 5-6 Gb of RAM with 40-60% CPU usage. Still get high frames even though the GPU is not running at 100% all the time like it used to. GPU fluctuates from as low as 1% back to 100%. I compared what AMD Adrenalin was telling me in the on screen display overlay with GPUZ and CPUZ and I'm pretty sure it's not a reporting error. Weird thing - I haven't had any crashes at all. Used previous Adrenalin version and most current with same results. Checked everything, and I mean everything. Used Superposition Benchmark and it shows it will go up to 11GB of VRAM and 5 or 6GB of RAM, used OCCT Perestroika which did show some GPU memory errors. Checked motherboard BIOS settings, even reflashed mobo BIOS, reflashed GPU bios using ATI Flash, I really think I am going to have to RMA this thing. Built this PC in July brand new. Had the GPU for 1 1/2 months, so I'm just outside the Newegg warranty :/

The big question I have is why is the card not performing like it used to (I'm guessing obviously a failure somewhere) but I figured if it was a software issue it could be fixed. Don't think it would be running like this because I don't have the SSD anymore, just running on the 1TB WD Blue HDD now but I don't think a HDD would make your CPU usage jump from 10% to as much as 60% along with the GPU fluctuation....

Main issues:GPU-Z Sensor Log.txt
Using about 3x-6x more CPU % and much more RAM
Fluctuation in GPU Utilization % and using much less VRAM

uploaded a couple images / files.. The gpuz gif is running pubg with vsync off (100+FPS) using 5GB of VRAM and 6GB of RAM...

DxDiag9-16-2018.txt

GPU-Z Sensor Log.txt

gpuz.gif

Link to comment
https://linustechtips.com/topic/972777-help-gpu-troubleshooting-problems/
Share on other sites

Link to post
Share on other sites

Did another test, found out if I lower the page file size of windows virtual memory and run a game, the game crashes with a windows error window about not enough memory. I lowered the page file size from default 7911Mb to 2048Mb and ran PUBG - Game crashed in about 10 seconds. Soooooo yeaahhhhhhh....... I believe I will be on my way to RMA this thing and upgrade to a 1070Ti......

Insert Meme: BUT WHAT DOES IT MEAN??!! (Nightmare before Christmas)
 

Link to post
Share on other sites

You seem to be knowledgeable enough that I doubt this is the problem, but I will ask anyways. Are you sure you have the proper GPU drivers installed? Don't go off of what Windows installs automatically. Windows doesn't know what the hell its doing and can cause issues like this. Another possible issue would be Windows (being its a fresh install) is updating in the background causing CPU usage to spike which can affect game performance. Check to see what all is hogging the CPU while gaming.

QUOTE ME IF YOU WANT A REPLY!

 

PC #1

Ryzen 7 3700x@4.4ghz (All core) | MSI X470 Gaming Pro Carbon | Crucial Ballistix 2x16gb (OC 3600mhz)

MSI GTX 1080 8gb | SoundBlaster ZXR | Corsair HX850

Samsung 960 256gb | Samsung 860 1gb | Samsung 850 500gb

HGST 4tb, HGST 2tb | Seagate 2tb | Seagate 2tb

Custom CPU/GPU water loop

 

PC #2

Ryzen 7 1700@3.8ghz (All core) | Aorus AX370 Gaming K5 | Vengeance LED 3200mhz 2x8gb

Sapphire R9 290x 4gb | Asus Xonar DS | Corsair RM650

Samsung 850 128gb | Intel 240gb | Seagate 2tb

Corsair H80iGT AIO

 

Laptop

Core i7 6700HQ | Samsung 2400mhz 2x8gb DDR4

GTX 1060M 3gb | FiiO E10k DAC

Samsung 950 256gb | Sandisk Ultra 2tb SSD

Link to post
Share on other sites

I have the newest AMD Adrenalin drivers and also used the previous software version with same results :/ 

 

Nothing has been updating or downloading in the background, i even used Intel driver and support assistant to check for any needed driver and software updates, which shows none. All the mobo software is updated - BIOS, Chipset, drivers etc. VRAM, RAM and CPU usage go back to normal when not running any games. Although i will say its weird the CPUZ shows the CPU usage multiplier to be fluctuating from 8-43 on all cores, im assuming because of power saving features. When in use the cores go to 41/42 like normal. Did a windows memory / check disk full scan which showed no memory errors.

 

I honestly think something fried in the GPU. Strangest thing ive seen when the GPU only uses 700MB VRAM on a game and the RAM is at 4/5 GB... Pretty confident in my original assessment of "gonna have to RMA this thing". I guess I am just asking for any additional info on this; if anyone has seen similar issues like this in a GPU and what was their experience on a solution, whether it be RMA and replacement or refurbished the card. Curious about what happened so i can prevent it from happening again.

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×