General Discussions

[Nickamus]

Hello everyone!

Sorry for the long post and incorrect format (if there is one to follow) for a first post! I would have sent AMD a support ticket, but I get this error: Link to Error. I'll try again another day.

My full system can be found below or here: My Current System

• CPU: AMD Ryzen 7 3800X
  
• Cooler: Corsair H100i RGB PLATINUM
• Motherboard: Asus PRIME B450M-K
• RAM: Corsair Vengeance RGB Pro 16GB (2x 8GB) DDR4-3200 CL16
• Storage 1: Crucial 525GB MX300 SATA M.2 SSD
• Storage 2: Western Digital Blue 1TB 2.5" SSD
• GPU: PowerColor Red Dragon Radeon RX 5700 XT
• Case: Corsair Crystal 280X RGB
• PSU: Corsair RM750x 
• Monitor: Asus TUF GAMING VG249Q
• OS: Windows 10 Pro (64-bit)

In short, my issue is a black screen, all input becomes non-responsive, sound still continues as normal, then finally, ends with a forced restart with no warnings or BSOD. When it boots back up, no minidumps were created, WhoCrashed shows nothing, and only events on Event Viewer were recorded (). Please note that this is only occurring, currently, only on Cyberpunk 2077. Currently, after a third or fourth full system wipe, formatted the SSD and re-installing Windows 10, I only get crash to desktops. Then the process usually happens again after a restart or something.

I have this issue since December 2020 and have no luck in fixing it. Anyone got a fix for this?

My current conclusions are:

1. AMD Ryzen 7 3800X and PowerColor Red Dragon Radeon RX 5700 XT are incompatible for Cyberpunk 2077
   • No issues when the pair was a 3800X and a nVidia 1660S
   • No issues when the pair was an Intel CPU 10th Generation i5 and a 5700 XT
2. Cyberpunk 2077 is the culprit and I'm losing the plot over my system for no reason

Based on videos and forums, the following temperatures seem to be normal for my chosen parts:

• CPU temperatures: Minimum (idle): 45°C, Maximum: 75°C
• GPU temperatures: Minimum (idle): 45°C, Maximum: 80°C and 100°C (junction)
  
• RAM temperatures: Minimum (idle): 30°C, Maximum: 55°C

Tests I've done

• Test 1: Turned off fast boot
• Test 2:
  • RAM back at base 2133MHz and CPU controlled by ASUS ["Normal" and not "Optimised"]
  • For the GPU, on the “Radeon Software”, I left it at “Automatic”
• Test 3: Same as Test 2 but for the GPU, I left it at “Undervolt”
• Test 4: Same as Test 2 but for the GPU, I left it at “Overclock”
• Test 5: Resat GPU and RAM, making sure all cables are plugged in and not loose, reseating and applying new thermal for the CPU
• Test 6: Tried the original PSU cables (I bought premium Corsair cables to rule them out)
• Test 7: Tried onboard GPU
• Test 8: Unplugged all unnecessary devices (only the keyboard & mouse connected)
• Test 9: Tried each RAM individually and different sockets individually
• Test 10: Reinstalled drivers and using "Recommended" and not "Optional"
• Test 11: Removed my Lego stand to stop the GPU drooping down
• Test 12: Used two dedicated 6+2 PCIe cables instead of one splitting into two 6+2 cable ends
• Test 13: Tried setting RAM at 3200MHz @ 16-18-18-36 1.2V
• Test 14: Tried MemTest86
• Test 15: Tried other games and see if the issue persists (when RAM is back at 3200MHz @ 16-18-18-36 1.35V)
• Test 16: Ran OCCT overnight
• Test 17: Removed an adapter for the GPU
• Test 18: Swapped GPU with my mate (he gave me his GeForce GTX 1660 Ventus XS 6G OC) to see if I get the same issue.
• Test 19: Added my GPU back to my system after I did a DDU with the nVidia card and a reinstall of Windows (again). Had to use diskpart because my third reinstall of Windows some how locked it from being reinstalled.
  
• Test 20: Recently, I've seen Cinebench being mentioned. Going to test both stress and stability test for both multicore and single core
• Test 21: Tried Ubuntu to rule out Windows doing something in the background
• Test 22: Tried a new motherboard
• Test 23: Tried a new PSU
• Test 24: Tried Prime95
• Test 25: Changed power options to "Ryzen Balanced"

Any other tests I should do? Can RAM be the culprit even if MemTest86 has no issues?

Results

• Tests 1 to 6, 8, 9, 11, 12, 14, 17, 25: Still crashes
• Test 7: Doesn't even output a signal - maybe it is a sign of a faulty motherboard?
• Test 10: Works fine, then after the next shutdown, still crashes
• Test 13: Can't even get out of boot
• Test 15: Other games don't cause the same issue
• Test 16: No errors running it but issue occurs when playing CP2077
• Test 18: No crashes for 3 days straight with 12 (day 1), 6 (day 2), and 4 (day 3) hours in each day with CP2077. Everything seems to work fine with a different card. My mate had zero issues with my card when he tested on their system (Intel CPU 10th Generation i5)
• Test 19: Crashes came back after one day. No issues with other games
• Test 20: No issues running it
• Test 21: Couldn't run the high intensive games I wanted so uninstalled it (badly uninstalled so had to use diskpart to clean it out)
• Test 22: Tried MSI MAG B550M MORTAR and still crashes
• Test 23: Tried Corsair RM750x and still crashes (originally a 650W of the same series)
• Test 24: Running Blend and Large FFTs were fine. But small and smallest FFTs kills my system (shutdown and not a restart) a second later - unless I tick the "Disable AVX2" box

[gavintek]

Hello, my system was having the same symptoms as yours with RX 5600 XT, except mine shows green screen instead of black screen. But i think that's because i am using HDMI cable, and happened in other games (Specifically red dead redemption 2, total war games and kingdom come deliverance).

I've been troubleshooting this since,.. forever?, and i saw a lot of RX 5600 and 5700 owners having similar issues with varying degrees of error circling around black/green screens. You have done a considerable amount of testing it seems, so you probably know this too.. Its a silicone lottery on a whole new level.

I wish i have the solution for you, but i am also at my wits' end on this one. Currently, i am doing similar thing like your test no. 18. My gtx 1660 super came yesterday and i am using it to see whether the problem persist. Yesterday, i played total war for hours with zero problem, but i will test it more (1 to 2 weeks) to really make sure that this is a GPU problem. If i have zero crashes after 1/2 weeks, i would RMA or sell my RX 5600 XT (should be easy to sell GPU these days as supplies are low and demands are high).

Things that might help you troubleshooting this:

- Can you check your event viewer, filter the events on you system with "18, 41". Every time you crash, do you see both event ID 18 and 41 in your event viewer?

- You can change your whocrashed settings to include WHEA error. These might capture the crash dump retrospectively.

[benman2785]

how did you clean the old Nvidia driver?

ps best would be a fresh Win10 20H2 install (as Nvidia does some stupid stuff to Windows)

[Nickamus]

---

---

@gavintek wrote:

Hello, my system was having the same symptoms as yours with RX 5600 XT, except mine shows green screen instead of black screen. But i think that's because i am using HDMI cable, and happened in other games (Specifically red dead redemption 2, total war games and kingdom come deliverance).

I've been troubleshooting this since,.. forever?, and i saw a lot of RX 5600 and 5700 owners having similar issues with varying degrees of error circling around black/green screens. You have done a considerable amount of testing it seems, so you probably know this too.. Its a silicone lottery on a whole new level.

I wish i have the solution for you, but i am also at my wits' end on this one. Currently, i am doing similar thing like your test no. 18. My gtx 1660 super came yesterday and i am using it to see whether the problem persist. Yesterday, i played total war for hours with zero problem, but i will test it more (1 to 2 weeks) to really make sure that this is a GPU problem. If i have zero crashes after 1/2 weeks, i would RMA or sell my RX 5600 XT (should be easy to sell GPU these days as supplies are low and demands are high).

Things that might help you troubleshooting this:

- Can you check your event viewer, filter the events on you system with "18, 41". Every time you crash, do you see both event ID 18 and 41 in your event viewer?

- You can change your whocrashed settings to include WHEA error. These might capture the crash dump retrospectively.

---

I eventually got "WhoCrashed" to point and find the WHEA errors.

When my computer black screens then restarts I get the following for the Event Viewer:

• Under Event Type, the "Error" event is always the "Event 18, WHEA-logger"
  • Note: The "Processor APIC ID" for the WHEA-logger error changes every time it occurs
• Under Event Type, the "Critical" event in Event Viewer, it's always the "Event 41, Kernel Power" message.

I got reply back from ASUS about my motherboard (I've already ruled out the motherboard, but wanted their opinion):

From the tests you've done one of the things that catches my eye is that Prime 95 forced your system to shut down, this definitely isn't normal and makes me think there's something not quite right.

You also mention that you've also used an MSI B550 motherboard, this would mean that the GPU would have been connected to the CPU via a PCIe 4.0 bus yet you still had the same issue.

The RX 5700 cards in particular suffered a poor reputation for crashes for certain users I don't believe that AMD were able to iron it out completely with driver updates, as you can see from the reply to your post others are still facing issues with first gen RDNA.

The only thing I could think of, other than this being yet another bug in Cyberpunk 2077, is that the CPU is is causing this crash, that would also explain the behaviour in Prime95, but I don't believe it's the motherboard.

I was going to try and go through Amazon for a replacement of my RAM, but if RAM was the issue, wouldn't it have at least one error in MemTest86 no matter which combination I tried?

I really believe the issue is to do with communication error or connection issues or something related to that. I based that off of how I fixed this issue back in September 2020 - removing one layer off my Lego GPU drooping/sagging stand - which led me to change the motherboard just in case the PCIe slot was a bust. But it is hard to say which the issue is since, like my test I've tried a AMD Ryzen 7 3800X + 1660S and an Intel 10th Gen i5 + 5700 XT both had no issues for 3 days. However, that could be a red herring and need at least months to test.

@benman2785 wrote:

how did you clean the old Nvidia driver?

ps best would be a fresh Win10 20H2 install (as Nvidia does some stupid stuff to Windows)

---

Well the first time was using DDU on Safe Mode. But it's been a week and a bit with 2 fresh window installs since then.

I am currently on the 20H2 with my 5700 XT since I had to give back my mate his 1660S.

[gavintek]

hmm, when your system shuts down on prime 95, do you get the same event ID like when you crashed while gaming (18 and 41)? and do you happen to test this when you had your 1660 installed? I just tested this just now on my ryzen 5 5600x and it did not shut my system down like yours. If you have your PBO turned on, maybe disable it and run it again / update your mobo bios and AMD chipset.

Its been 3 days since i am using gtx 1660 super, and so far zero crashes. But like you said, maybe a red herring. I also factored in the fact that i did not game that much this week due to work commitments.

Like yours, my WHEA shows different processor core each time it crashes. I used Windbg Preview (windows debugger) on my WHEA 18 error and found this:

MODULE_NAME: AuthenticAMD

IMAGE_NAME:  AuthenticAMD.sys

STACK_COMMAND:  .thread ; .cxr ; kb

FAILURE_BUCKET_ID:  LKD_0x124_0_AuthenticAMD_BANK0_MSCOD0000_MCACOD0108_PCC_UC_IMAGE_AuthenticAMD.sys

OSPLATFORM_TYPE:  x64

OSNAME:  Windows 10

FAILURE_ID_HASH:  {509acb9c-038f-dfdb-adc2-7917670271d1}

Followup:     MachineOwner

Heard somewhere that this issue only persist when playing games that are optimized for nvidia/intel (such as cyberpunk). Instruction set sent to CPU make the CPU access bad memory on GPU, causing cache hierarchy error, or something like that.

[delevero]

I also have crashes but i have a 5900x cpu and a 6900xt card. with other words the newest of the newest hardware.
For me it helped allot that i unplug my vr headset even though i play a non vr game then this effected my games and made them crash..

Well at least for a while. Now i experience crashes now and then with black screen and the computer totaly shut off.
Its is not a heating problem i tested this many times, and i can also run all kind of benchmarking and stress test software that test both cpu and gpu and the tests are rock steady.
I also tested my Ram via the "dos" ram test that take ages to complete and there is 0 issues.

I think at this point that it must be a driver problem and in my case i think its the graphics cards driver that make the games crash.
But i also think a little bit if it could be the last two bios updates that optimize the ryzen 5000 series cpu so they perform better.

All of my crashes started when i switch from the 2080rtx nvidia to the new 6900xt so my bet is the driver have a bug.
I imagine that the 5700 gpu that you talk about might share some of the same programming so it could be a bug that effect all amd graphics cards.

[benman2785]

@delevero 

how did you clean old Nvidia driver from system?

[delevero]

I found this news today that maybe is the reason why our computers crash or shut down.
There is a bios update on the way to all AMD motherboards
https://www.techpowerup.com/276833/amd-announces-agesa-1-1-9-0-firmware-updates-improve-fclk-oc-stab...

[Nickamus]

---

@gavintek wrote:

hmm, when your system shuts down on prime 95, do you get the same event ID like when you crashed while gaming (18 and 41)? and do you happen to test this when you had your 1660 installed? I just tested this just now on my ryzen 5 5600x and it did not shut my system down like yours. If you have your PBO turned on, maybe disable it and run it again / update your mobo bios and AMD chipset.

Its been 3 days since i am using gtx 1660 super, and so far zero crashes. But like you said, maybe a red herring. I also factored in the fact that i did not game that much this week due to work commitments.

Like yours, my WHEA shows different processor core each time it crashes. I used Windbg Preview (windows debugger) on my WHEA 18 error and found this:

MODULE_NAME: AuthenticAMD

IMAGE_NAME:  AuthenticAMD.sys

STACK_COMMAND:  .thread ; .cxr ; kb

FAILURE_BUCKET_ID:  LKD_0x124_0_AuthenticAMD_BANK0_MSCOD0000_MCACOD0108_PCC_UC_IMAGE_AuthenticAMD.sys

OSPLATFORM_TYPE:  x64

OSNAME:  Windows 10

FAILURE_ID_HASH:  {509acb9c-038f-dfdb-adc2-7917670271d1}

Followup:     MachineOwner

Heard somewhere that this issue only persist when playing games that are optimized for nvidia/intel (such as cyberpunk). Instruction set sent to CPU make the CPU access bad memory on GPU, causing cache hierarchy error, or something like that.

---

Sadly, I do not have the 1660S to test anymore. The only test I did with it at the time was let CP2077 run 3 days in a row with no issues (and my mate did the same with his setup - maybe less hours than me).

Currently using the latest BIOS (it's a beta according to Asus) and crashed on the second day. I haven't actually tried downgrading the BIOS to see maybe the newer updates are the cause of the issues. Only problem is I can't remember what the BIOS version was installed back in August 2020 and I have to be careful to not do downgrade too far and lose the compatibility of my CPU haha.

I think I tried turning off the CPU boost thing in the BIOS but can't actually remember.

[Nickamus]

Tested some games today and yesterday. 

Few hours on Path of Exile had zero issues today and yesterday, only disconnects and crashes to desktop - might downgrade to the recommended driver rather than the optional recent one.

Yesterday, played Age of Empires II: Definitive Edition no FPS lag. Today, started the game up and 2 minutes later, black screen, sound kept going and the restart. Once restarted, the game didn't crash, but the game performance was bad (at the lowest graphic settings I was getting between 1 to 50 FPS).