Asus Strix-RX470-4g-gaming blackscreens then no signals

[fiveofeight]

Computer Type: Selfbuilt desktop
GPU: Asus Strix-RX470-4g-gaming with Hynix vram. Stock frequencies and power curve.
CPU: Intel Core i7 4770k stock frequencies
Motherboard: ASRock Z97 Killer Fatal1ty
RAM: 2x 8GB Gskill F3-12800CL10D-16GBXL DDR3 ram 1600MHz 10:10:10:30
PSU: Corsair GS700
Operating System & Version: Windows 10 creator's update
GPU Drivers: Crimson 17.6.1
Background Applications: Any sort of 3d game such as Overwatch, WoW or Unigine Valley Benchmark

Description of Original Problem: After playing games for a few hours, my screen will go blank and then both my monitors say no signal (DVI an DP to VGA adapter are the two connectivity ports I use). I can still hear audio playing from whatever game I was playing, VOIP call I'm in or music I was listing to for around 30 seconds after the crash, after that it goes quiet. I have to hard reset the machine to get it to respond again. After rebooting sometimes event log will say "Application Wow-64.exe has been blocked from accessing Graphics hardware. The card does not appear to be overheating, hottest it ever gets is 72c and it seems to happen mostly when a game is running, I've tested mining Ethereum (with Claymore) and the system doesn't black screen like when playing a game. I have also tested just using the DVI monitor rather than both to make sure the DP-VGA adapter wasn't faulty.

Troubleshooting: I've put in my old MSI Radeon HD 5770 and used DDU to clean up the drivers, the issue does not present itself with the old card. I've used DDU to try different driver versions with the Rx 470, tried 16.8.3, 16.10.2, 16.11.5, 16.12.2, same result. Used the registry change to change the TDR to 8, no change. I tried dropping the clocks by 300MHz (1206 to 906 core and 1650MHz to 1350MHz for ram), and giving the card 15% more power, neither made any difference. I've tried putting it in a different PCI-Express slot on my main PC. I've put the card in another machine (Intel Pentium G 3220, Asrock H81M motherboard 700W thermaltake PSU, Windows 7 Home) and experienced the same black screen issue (was using 16.50 driver). I also tried Win+Ctrl+Shift+B to reset the graphics driver after I get a black screen, but this did nothing.

My display port ended up failing and I submitted an RMA to Asus with both issues listed and got the card replaced. The DP port works on the replacement card but the no signal crash issue is still present same as before.

[Bill]

Hello fiveofeight
Did you find out why your system had blocked your game from accessing Graphics hardware?
Any error message or code regarding that matter?
Do you have the full description on related events or logs?
Have you tried using the latest driver available on AMD website which is the Crimson ReLive Edition 17.7.2?
Thank you

[fiveofeight]

Hi Bill, thanks for the reply.

Don't have any logs or errors aside from what shows up in event viewer. Wow doesn't have any history of the crash, neither does Unigine. I don't think the amd drivers store any logs outside of event viewer.

I have not figured out what is causing it. I reset TDR from 8 seconds to 2, and with today's crash instead of getting "Application Valley.exe has been blocked from accessing Graphics hardware.", I got "Display driver amdkmdap stopped responding and has successfully recovered. The computer never recovered however, the screen said no signal for around 10 minutes before I pulled the power to reset the machine. I tried to remote desktop into the machine from my laptop, and it timed out. It also seems like USB stops working when the GPU crashes, my keyboard stops working so I can't see if win+ctrl+shift+b resets the driver (I'm pretty sure it doesn't work cause my volume buttons do nothing, and after I unplug and replug the keyboard, the lights don't turn back on till the computer is restarted.)

Some things still work while the computer is in this state however. I use Voicemeeter Banana to send audio from my laptop to my desktop PC, and audio kept playing for the 5 minutes I left the PC running after it crashed. The event log also keeps logging events after the display crash, but it only shows generic stuff like: "[DistributedCOM]: The server Microsoft.Windows.Photos_2017.18062.13720.0_x64__8wekyb3d8bbwe!App.AppXy9rh3t8m2jfpvhhxp6y2ksgeq77vymbq.mca did not register with DCOM within the required timeout." The only other errors that appear are the expected "The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly." because I had to hard power off the PC.

Went through some of my older logs, and found on June 16'th "Application Wow-64.exe has been blocked from accessing Graphics hardware." immediately followed by "Display driver amdkmdap stopped responding and has successfully recovered.".

I tried installing Radeon Crimson 17.7.2 a few days ago (picked the clean install option when I upgrade, but I didn't use DDU to uninstall the old drivers), the computer ran longer, but crashed yet again today after around 100 hours. Going to try using DDU now and retesting with Unigine Haven and Crimson 17.7.2.

While the Unigine haven benchmark is running, temperature doesn't go above 76c, so it doesn't seem like it's an overheating issue.

[Bahz]

You may have the older vBIOS, there's a new vBIOS released that fixes this issue.
http://dlcdnet.asus.com/pub/ASUS/vga/ATI/STRIX_RX470_OC.rar?_ga=2.193249773.1140056051.1501426809-44...

[fiveofeight]

Hi Bahz,

Is that vBIOS for the Strix-RX470-o4g-gaming or the Strix-RX470-4g-gaming (I have the non-oc card). I ask since the file linked has OC in the name. I tried flashing my card with your link and it said: Your graphic card no need to update VBIOS!

[fiveofeight]

Uninstalled the drivers and installed 17.7.2, after around 100 hours the computer black screen and no signaled me again. Though this time audio stopped playing immediately. Checked the eventlog afterwards and found out the computer BSOD'd on me:
The computer has rebooted from a bugcheck. The bugcheck was: 0x0000003b (0x00000000c0000005, 0xfffff80390294a64, 0xffff8080a0fa8b30, 0x0000000000000000). A dump was saved in: C:\WINDOWS\MEMORY.DMP. Report Id: 2e539412-3c7a-4d18-ba06-988e48dc6ef4.

I uploaded the crashdump here: https://www.dropbox.com/s/qhsdypmamf119l0/080517-5828-01.dmp?dl=0

[Bahz]

fiveofeight wrote:
Uninstalled the drivers and installed 17.7.2, after around 100 hours the computer black screen and no signaled me again. Though this time audio stopped playing immediately. Checked the eventlog afterwards and found out the computer BSOD'd on me:
The computer has rebooted from a bugcheck. The bugcheck was: 0x0000003b (0x00000000c0000005, 0xfffff80390294a64, 0xffff8080a0fa8b30, 0x0000000000000000). A dump was saved in: C:\WINDOWS\MEMORY.DMP. Report Id: 2e539412-3c7a-4d18-ba06-988e48dc6ef4.

I uploaded the crashdump here: https://www.dropbox.com/s/qhsdypmamf119l0/080517-5828-01.dmp?dl=0

Looks like you have a defective card, probably need to RMA it.

[fiveofeight]

I've tried to contact Asus to arrange another RMA, but they've ignored the 3 mails I've sent. I messaged them on July 24 (N170703656), July 28, and August 7'th.

[Bahz]

fiveofeight wrote:
I've tried to contact Asus to arrange another RMA, but they've ignored the 3 mails I've sent. I messaged them on July 24 (N170703656), July 28, and August 7'th.

Can you send me a PM with the following information?

S/N:
Full name:
Contact #:
Email:
Case ID#s:
Previous RMA#s (if there is):

Thanks.

[fiveofeight]

So, update to this situation. I ended up receiving no response to any of the 3 tickets that I submitted. I called Asus after waiting another two weeks, and got an RMA issued through the phone. I asked the agent about why I never received a response to any of my tickets and provided him with the Case ID#. He looked it up and saw that Asus had received them, but had no explanation for why they never replied to me.

I mail the card back, and receive the replacement, and the issue persists. My testing got postponed midway through thanks to Hurricane Maria taking down the power grid (I live in Puerto Rico, US), and resumed testing after it got restored. The same issue persists, I get no signal/black screen crashes when the GPU is installed in the PC and its been running for a few hours to a few days (never can get past 14 days of uptime, most often it's less than a week). Ran the main PC (4770k) using an iGPU and got 20 days of uptime on it, so without the RX 470 its stable. It looks to be the same issue as described in the RX 470 OC4GB thread: https://rog.asus.com/forum/showthread.php?93691-ASUS-ROG-Radeon-RX-470-Strix-Gaming-OC-4GB-loses-con... But the non-oc 470 has received no such BIOS update, and it seems that bios did not fix the issue for the card anyway.

I replied to the last ticket I have with Asus explaining the issue is still persistent after receiving 2 RX 470s with the same issue, and the shipping cost is mounting up. I asked if I could get the card replaced with a different model (such as an equivalent NVIDIA card) or they cover the shipping both ways, and the request got ignored, he simply told me to submit yet another RMA. I do not want to send back the card to have the same model with the same issue persistent, so I need a different solution to be found for this issue.

I've documented the whole thing over on a reddit post here: https://www.reddit.com/r/ASUS/comments/7dnetc/terrible_rma_experience_with_asus_rx_470/