Windows 10: Bluescreen and forced restarts on new ryzen system

Hello,


I have built a brand new system recently:


Mobo: MSI MPG x570 Gaming Plus

CPU: Ryzen 3700x

GPU: RTX 2070 Super

RAM: tried 2 different kits, Corsair Vengeance LPX 2x8GB DDR4 3200MHz CL 16 and Patriot Viper 4 2x8GB DDR4 3200MHz CL16

PSU: bequiet Straight Power 11 550W

in Corsair Carbide R270 case


The only thing that is not brand-new is the formatted SSD and a HDD with backup-files on it (mostly images. I have not installed nor run any programs from that drive yet). Error Checks via right-clicking the drives in the "This PC" tab and navigating to tools -> Check reported no errors.


Fresh windows install via bootable external harddrive. First thing I did was M-Flash the Bios to the current version, then installed all drivers from the respective websites (chipset including audio, lan etc), GPU, followed by Anti-Virus software and then others (Logitech drivers, MSI Afterburner, CPU-Z, etc).


I don't quite remember when it first started, but very early into that process the system started bluescreening. I installed all windows updates after it started happening.


These are the minidumps: https://1drv.ms/u/s!AtBvJqmo747yhi7VxJX4LcU_v9bG?e=XvlvXH


There are only 5 of them - there have been many more crashes. Several times the system bluescreened without displaying any error, just stating that it was investigating the issue, then I could restart. This screen did not go away, forcing me to press the reset button. This happened about as often as normal bluescreens with the attached minidumps.


Instead of running a memtest for hours on end I decided to swap out the RAM, which did not help. I thought it did, the system didn't crash for several hours but then it happened again.


I noticed that the crashes tend to happen in quick succession, followed by a period of no crashes.


I ran several Cinebench R20 benchmark with no crashes and played quite a bit of Witcher 3 with no crashes either.

CPU temps were quite high during the bench - 90° - but did not cause a crash. GPU temps have been very low even under load, sub 70. The case feels very cool inside.


What could be causing these intermittent bluescreens? I tried analyzing the minidumps via windbg but that did not clear things up for me at all.


I'd very much appreciate any help and suggestions. I don't know what the issue could be.

:)

 

AMD Ryzen 2200G systems hang, usually no BSOD

Problem: Systems hard lock after anywhere from a few minutes of uptime to a couple of days. Image is frozen on screen and no mouse cursor is visible, pressing Ctrl+Shift+Win+B to reset graphics driver does nothing. Only one of the PCs has ever bluescreened
and it only did so one time, with error code 0x1000009f DRIVER_POWER_STATE_FAILURE for ntoskrnl.exe. I tried updating, downgrading, and reinstalling GPU, NIC, and sound drivers but no luck. Every other time the PCs don't bluescreen when they lock up and there
is nothing in the Event Viewer that is helpful.

Troubleshooting steps tried:

1. Updated AMD drivers from 18.5.1 to 18.7.1 and 18.8.1.
2. Clean installed both 18.8.1 and 18.5.1.
3. Installed Realtek LAN and audio drivers from ASRock.
4. Clean installed 17.7 AIO driver package from ASRock (listed as 17.40).
5. Updated BIOS from 4.70 to 4.90.
6. Reset BIOS.
7. Uninstalled August 2018 Windows 10 patches.
8. Switched to Balanced power plan instead of Ryzen Balanced.
9. Disabled USB suspending and PCIe ASPM
10. Switched to High performance, disabled ASPM & USB SS
11. Increased SOC voltage to 1.1 and RAM voltage to 1.3.
12. Removed AMD drivers, used Microsoft Basic Display Adapter.
13. Uninstalled Malwarebytes Premium 3.5.1
14. Started in Safe Mode w/ Networking
15. Repair install 1803.
16. Used only one RAM stick at 2133MHz.

Next steps:

1. Clean install 1803
2. Install 1709, delay 1803 update
3. RMA ASRock motherboards
4. Replaced ASRock motherboards
5. Replace AMD with Intel

Identical systems specs:

• AMD Ryzen 2200G CPU
• ASRock AB350M Pro4 Micro ATX Motherboard
• G.Skill Ripjaws V 8GB (4GBx2) 2666MHz DDR4 RAM
• Samsung 860 EVO 250GB SSD
• EVGA 450W Bronze PSU
• Fractal Design Core 1000 Micro ATX Case
• Windows 10 Pro x64 v1803

Other system specs:

• AMD Ryzen 2200G CPU
• Asus TUF B350M Plus Gaming Micro ATX Motherboard
• Corsair Vengeance LPX 8GB (4GBx2) 3000MHz DDR4 RAM
• Samsung 860 EVO 250GB SSD
• Corsair TX650M Gold PSU
• Fractal Design Node 804 Micro ATX Case
• Windows 10 Pro x64 v1803

Details:

I built four identical systems and one with different motherboard and RAM, all of them are having this problem to varying degrees of severity. The worst one locked up within 15 minutes, but the best one used to work for over a week without issues. I tried
doing a repair install on one of the systems, during the preparation period for the upgrade the system locked up at 16%. I was able to get it to complete to the point of restarting to do the majority of the install, but that locked up multiple times as well.

I've looked through the event log, and there is nothing between the systems that points to the source of the problem. There is no one thing that causes the systems to lockup, and there is no rime or reason to when they lockup. One of them did seem to lockup
every hour, but otherwise it happens at random.

I'm at my wits end; in over 15 years of troubleshooting and building PCs I have never had issues of this scope or inconsistency. I repair computers for a living, and I have never seen issues like this. I have a server based on AMD's Epyc 7351P and my main
desktop is based on the Ryzen 7 1700, and I haven't had issues anywhere close to this on those systems. At this point I'm seriously considering replacing the CPUs and motherboards with Intel equivalents. I've I'm willing to try any suggestions any one has,
no matter how much of a long shot they are or how stupid they seem.

Here's a link to the dump file, driver list, and system info: https://1drv.ms/f/s!AoFuBFiKuAH5uWqmul1JHUSusJ4S

 

Windows Updated(16299.309).KB4088776 Restart is required to complete.


Hi dalchina. Good news, I followed your suggestion and reinstalled the update and now it appears as "installed correctly" *Biggrin Maybe it was not necessary to run the Windows Update Troubleshooter or maybe if necessary. But using it I do not think it could have worsened something. The installation was very fast, unlike the first time it showed step 1 and step 2 completed.

Thanks for the suggestions and for the time to respond Best regards.


And thanks for the tip of that program, I'll have it on my list.

 

forced restarts

it has four 12V rails rated at 19A. thats approx 220W per rail. My previous 600W PSU had four 18A rails and it was listed as a 600W.


Since the PSU has multpile rails instead of one single one, its entirely possible say, for your video card and CPU to end up on the same rail - and with an i7 and a modern video card, 220W isnt going to be enough.