Question R9 290 Sapphire Tri-X, "Display driver stopped responding" crash

Jun 13, 2019
2
0
10
0
I don't have the error log as I removed the OS partition that I was using to test on. But I am in need of some help on diagnosing this on-going issue.

Problem:
Card was running fine until Windows 10 rolled around and this is when all hectic started to happen even until this day. Drivers seem to crash after every game. I put the card on the shelf and today I am wanting to revive the card and see what I can do. Here's a list of whats going on.
A) The GPU loads normally. B) Display functions in-game only until you try to exit the game is when the display driver stops responding and force restart the entire PC. C) Temperature on the R9 290 Sapphire while gaming and under load is at 60 degrees C, so I don't think it is overheating. D) Before partition the OS where I was running the test, the error log reads: Livekernalevent critical issue.

When running the GPU on Windows 7 OS, playing Apex Legends, an error would sometimes appear that reads, "Cannot write to memory error," and cause the PC to force restart again.

My Check-off list of things I've done to narrow down the issue:
Reinstalling Windows 10 OS with fresh state to make sure it is not corrupted Windows file.
Performed an integrity check on Windows file with DISM and scf /scannow.
Changed registry values for TDR (Time-out and recovery) to delay to give GPU some more time to recover itself.
Tested the GPU on Windows 7 to see if there's a change of behavior and ruling out if Windows 10 is the culprit.
Increased the powerslider on the AMD software to max, set power management on PC to performance to rule out insufficient power.
Turned off Anti-alias as some users had luck with Wattman crashing to stop with this.
Entered BIOS and gave priority to PCIe and enabled duo GPU + Integrated graphic to rule out possible motherboard not being able to read the dedicated GPU.
I removed the Sapphire cooler and inspected the PCB to make sure there are no obvious signs of failing parts such as burnt marks, capacitors missing, etc...
Cleaned the PCB where I believe too look like "can" interfere with electricity flow with alcohol and replaced the thermal paste with a better one.
Uninstalled AMD drivers with DDU and tried running with generic drivers and then testing it with an older driver before the problem started happening to rule out driver issues.
Did a complete uninstall of AMD drivers with DDU in safe mode and only ran custom install for the display driver itself to rule out some sort of software issue.

I am running out of ideas how how to repair this issue:
I have not done a memory test on the GPU (don't know how), but I am leaning to hardware failure. And if it is hardware failure, how can I pinpoint the part on the PCB that is failing and what is causing the failure and what I can do to replace those parts. If anyone have any idea on what might be the problem, please shed some light.

Things to note:
Running it on Windows 10, it doesn't show up as "Display driver stopped responding." Instead, it crashes and "Wattman has been restored due to system failure." And sometimes it just force restart the PC. Only on Windows 7 does the error report as, "Display driver stopped responding." Also to note, for odd reasons, when using HDMI port... the boot screen doesn't display properly only until the Windows log-in appears does the display give signal. Using the DVI port, boot screen displays properly.

System Specs:
Windows 10
DDR3 two memory sticks, 4GB each.
1TB HDD
R9 290 Sapphire Tri-X on Uber-switch mode.
Corsair 600M CX
Asus Q87M-E/CSM
 
Last edited:
Jun 13, 2019
2
0
10
0
I realize yesterday that the Sapphire R9 290 Tri-X comes by default Overclocked. I started to Underclock the card to troubleshoot and narrow down the problem.

The more underclock I did for the memory clock, lessen the hang time and artifacts visual when the GPU crashes. Underclocking the core to 500 mhz from 1000 mhz and the memory to 1000 mhz from 1300 mhz has stopped the PC from overly aggressive restarting, but the hang and artifact visual still occurs only after leaving/exiting a game.

Increasing visual or high quality in-game such as Antialias causes the GPU to go into crashes more "quicker" than usual. My thoughts are pointing to the GPU VRAM. But something still bothers me. If this was the case, I would of seen an immediate reaction with hang and crashes during game while VRAM is being used. The crashes/BSOD/hang time happens only after I exit the game.

Also, it seems to be a universal problem occurring on all GPU just by looking at the forums. It seems as if most people are constantly having their GPU crash on them frequently, even newer GPU.
---

If anyone have a clue or possible fix to the issue please share.
 

ASK THE COMMUNITY