Weird PC lockups

sygnus21 · 18 May 2020

Hey guy/gals

Recently I've been having random issues where the entire PC will just hard lock. The only way out is the power or reset switch. It seems to only happen while I'm on the internet with multiple tabs open. Doesn't matter the site or page, but I do notice the longer I'm on the net the more likely the issue will occur. Outside of that I can game for hours, play music, or do photo edits without issue.

Checking Reliability history shows the system suffered a "hardware" error. Current system specs are listed (system One). System is not overclocked. XMP for RAM is enabled.

Thanks.

zbook · 18 May 2020

Hi sygnus21,

The computer has been having problems related to the GPU.

The logs displayed BSOD since 5/2/2020.

Live kernel events related to the GPU were recorded in February.

The bugchecks were 116: Video TDR failure

Replace the GPU drivers > see if there is an impact on computer stability / instability.

If automatic restart is checked it can be easy to miss BSOD.
So settings will be modified to make it easier to see BSOD.

1) Open administrative command prompt and type or copy and paste:
2) sfc /scannow
3) dism /online /cleanup-image /scanhealth
4) dism /online /cleanup-image /restorehealth
5) sfc /scannow
6) chkdsk /scan
7) wmic recoveros set autoreboot = false
8) wmic recoveros set DebugInfoType = 7
9) wmic recoveros get autoreboot
10) wmic recoveros get DebugInfoType
11) bcdedit /enum {badmemory}

12) When these have completed > right click on the top bar or title bar of the administrative command prompt box > left click on edit then select all > right click on the top bar again > left click on edit then copy > paste into the thread

13) Make sure that there is no over clocking while troubleshooting

14) In the left lower corner search type: system or system control > open system control panel > on the left pane click advanced system settings

a) > on the advanced tab under startup and recovery > click settings > post an image of the startup and recovery window into the thread

b) > on the advanced tab under performance > click on settings > on the performance options window > click on the advanced tab > under virtual memory > click on change > post an image of the virtual memory window into the thread

15) Choose one of the temperature monitoring software applications: Speecy, HW monitor, Speed fan:
Speccy - Free Download - Piriform: Speccy - System Information - Free
Download Speccy | Find your computer specs, free!
Download Speccy | Find your computer specs, free!
HWMONITOR | Softwares | CPUID: HWMONITOR | Softwares | CPUID
HWMONITOR | Softwares | CPUID
SpeedFan - Access temperature sensor in your computer: SpeedFan - Access temperature sensor in your computer
SpeedFan - Access temperature sensor in your computer

16) Uninstall everything AMD using Display driver uninstaller and install new drivers from the AMD website.

16) On the AMD web site be sure the "clean install" box is checked and only install the graphics driver.

Official Display Driver Uninstaller DDU Download

Display Driver Uninstaller Download version 18.0.2.5 (or newer version if available)

Display Driver Uninstaller: How to use - Windows 7 Help Forums

http://support.amd.com/en-us

https://www.amd.com/en/support/kb/faq/rs-install

Code:

Name	AMD Radeon VII
PNP Device ID	PCI\VEN_1002&DEV_66AF&SUBSYS_081E1002&REV_C1\6&26D0B864&0&00000008
Adapter Type	AMD Radeon Graphics Processor (0x66AF), Advanced Micro Devices, Inc. compatible
Adapter Description	AMD Radeon VII
Adapter RAM	(1,048,576) bytes
Installed Drivers	C:\Windows\System32\DriverStore\FileRepository\u0352369.inf_amd64_275818c45a90b362\B352355\aticfx64.dll,C:\Windows\System32\DriverStore\FileRepository\u0352369.inf_amd64_275818c45a90b362\B352355\aticfx64.dll,C:\Windows\System32\DriverStore\FileRepository\u0352369.inf_amd64_275818c45a90b362\B352355\aticfx64.dll,C:\Windows\System32\DriverStore\FileRepository\u0352369.inf_amd64_275818c45a90b362\B352355\amdxc64.dll
Driver Version	26.20.15019.19000
INF File	oem71.inf (ati2mtag_Vega20 section)
Color Planes	Not Available
Color Table Entries	4294967296
Resolution	1920 x 1200 x 60 hertz
Bits/Pixel	32
Memory Address	0x70000000-0x801FFFFF
Memory Address	0x80000000-0x801FFFFF
I/O Port	0x00003F00-0x00003FFF
Memory Address	0x9AA00000-0x9AAFFFFF
IRQ Channel	IRQ 4294967208
Driver	C:\WINDOWS\SYSTEM32\DRIVERSTORE\FILEREPOSITORY\U0352369.INF_AMD64_275818C45A90B362\B352355\ATIKMPAG.SYS (26.20.15019.19000, 576.02 KB (589,840 bytes), 3/20/2020 00:29)

Code:

atikmpag.sys Fri Feb 28 17:11:58 2020 (5E599E3E)

zbook · 18 May 2020

Hi sygnus21,

The computer has been having problems related to the GPU.

The logs displayed BSOD since 5/2/2020.

Live kernel events related to the GPU were recorded in February.

The bugchecks were 116: Video TDR failure

Replace the GPU drivers > see if there is an impact on computer stability / instability.

If automatic restart is checked it can be easy to miss BSOD.
So settings will be modified to make it easier to see BSOD.

1) Open administrative command prompt and type or copy and paste:
2) sfc /scannow
3) dism /online /cleanup-image /scanhealth
4) dism /online /cleanup-image /restorehealth
5) sfc /scannow
6) chkdsk /scan
7) wmic recoveros set autoreboot = false
8) wmic recoveros set DebugInfoType = 7
9) wmic recoveros get autoreboot
10) wmic recoveros get DebugInfoType
11) bcdedit /enum {badmemory}

12) When these have completed > right click on the top bar or title bar of the administrative command prompt box > left click on edit then select all > right click on the top bar again > left click on edit then copy > paste into the thread

13) Make sure that there is no over clocking while troubleshooting

14) In the left lower corner search type: system or system control > open system control panel > on the left pane click advanced system settings

a) > on the advanced tab under startup and recovery > click settings > post an image of the startup and recovery window into the thread

b) > on the advanced tab under performance > click on settings > on the performance options window > click on the advanced tab > under virtual memory > click on change > post an image of the virtual memory window into the thread

15) Choose one of the temperature monitoring software applications: Speecy, HW monitor, Speed fan:
Speccy - Free Download - Piriform: Speccy - System Information - Free
Download Speccy | Find your computer specs, free!
Download Speccy | Find your computer specs, free!
HWMONITOR | Softwares | CPUID: HWMONITOR | Softwares | CPUID
HWMONITOR | Softwares | CPUID
SpeedFan - Access temperature sensor in your computer: SpeedFan - Access temperature sensor in your computer
SpeedFan - Access temperature sensor in your computer

16) Uninstall everything AMD using Display driver uninstaller and install new drivers from the AMD website.

17) On the AMD web site be sure the "clean install" box is checked and only install the graphics driver.

Official Display Driver Uninstaller DDU Download

Display Driver Uninstaller Download version 18.0.2.5 (or newer version if available)

Display Driver Uninstaller: How to use - Windows 7 Help Forums

http://support.amd.com/en-us

https://www.amd.com/en/support/kb/faq/rs-install

Code:

Name	AMD Radeon VII
PNP Device ID	PCI\VEN_1002&DEV_66AF&SUBSYS_081E1002&REV_C1\6&26D0B864&0&00000008
Adapter Type	AMD Radeon Graphics Processor (0x66AF), Advanced Micro Devices, Inc. compatible
Adapter Description	AMD Radeon VII
Adapter RAM	(1,048,576) bytes
Installed Drivers	C:\Windows\System32\DriverStore\FileRepository\u0352369.inf_amd64_275818c45a90b362\B352355\aticfx64.dll,C:\Windows\System32\DriverStore\FileRepository\u0352369.inf_amd64_275818c45a90b362\B352355\aticfx64.dll,C:\Windows\System32\DriverStore\FileRepository\u0352369.inf_amd64_275818c45a90b362\B352355\aticfx64.dll,C:\Windows\System32\DriverStore\FileRepository\u0352369.inf_amd64_275818c45a90b362\B352355\amdxc64.dll
Driver Version	26.20.15019.19000
INF File	oem71.inf (ati2mtag_Vega20 section)
Color Planes	Not Available
Color Table Entries	4294967296
Resolution	1920 x 1200 x 60 hertz
Bits/Pixel	32
Memory Address	0x70000000-0x801FFFFF
Memory Address	0x80000000-0x801FFFFF
I/O Port	0x00003F00-0x00003FFF
Memory Address	0x9AA00000-0x9AAFFFFF
IRQ Channel	IRQ 4294967208
Driver	C:\WINDOWS\SYSTEM32\DRIVERSTORE\FILEREPOSITORY\U0352369.INF_AMD64_275818C45A90B362\B352355\ATIKMPAG.SYS (26.20.15019.19000, 576.02 KB (589,840 bytes), 3/20/2020 00:29)

Code:

atikmpag.sys Fri Feb 28 17:11:58 2020 (5E599E3E)

jdc1 · 18 May 2020

I suggest disabling XMP while we troubleshoot. Although it's not likely to cause issues, I've seen it happen in the past.
Feel free to turn it back on once we're finished troubleshooting.

Please check and ensure that you have ALL available Windows Updates

You have 2 memory dumps (BSOD's) in the uploaded reports (15 May and 18 May)
Running them in my debugger as I type this.............
But, from the Error Reports folder, the most likely cause is video
The MSINFO32 report shows that the BSOD's mentioned above were both STOP 0x116 errors (video)

Your BIOS/UEFI version F9c dates from December of 2019
Please check the Gigabyte support website for your system/motherboard to see if there are any updates available. If there are, please update the BIOS/UEFI immediately.

You have 5 drives installed. I use this as a rough gauge of how much equipment is installed on the system. With this many drives, it's also more likely that you have other hardware installed. And this raises the question - "Is your PSU able to handle this amount of hardware - both during normal operating conditions and under severe "surge" conditions. The EVGA PSU in your spcs (1000 watt) is probably a good PSU and it likely has enough wattage to handle your extra equipment.

As stated earlier, the BSOD's are both STOP 0x116 errors.
More info on that here: Video TDR Troubleshooting
The article is a bit old (and some of the links are broken), but it's still relevant to your situation.
Basically, what it says is:
- perform an uninstall of the video drivers, then use DDU to remove any traces (free from here: Display Driver Uninstaller (DDU) | Wagnardsoft )
- download and install the latest available video drivers from AMD ( https://www.amd.com/en/support/kb/faq/gpu-56 )
Test to see if this helps

If not, then check:
- that the fan/cooler is clean and free of dirt, dust or hairballs
- that the card is firmly seated into the motherboard, and that any other cables are firmly seated into the card.
- that the slot that the card is in is free of dirt/dust/hairballs (blow it out with canned air - DO NOT USE A VACUUM OR AIR COMPRESSOR!!!)
- that the fan isn't sticking or sluggish (found by observing the card during operation. If it's sticking, when it sticks it'll cause problems.

Acronis drivers date from 2017. Please ensure that you have the latest updated drivers for your version of Acronis.
FYI - I own multiple copies of Acronis and only use the bootable version to make images. As such I don't need the Acronis software installed on my system - so I uninstalled it after making the bootable USB drive(s).

The following drivers date from 2010 and should be uninstalled:

Code:

SiLib.sys                                     Thu Jul 15 18:08:26 2010 (4C3F86DA)
USBXpress Development Kit http://www.silabs.com/SUPPORT/Pages/default.aspx 

SiUSBXp.sys                                   Thu Jul 15 18:27:13 2010 (4C3F8B41)
Silicone Labs USBXpress Development Kit http://www.silabs.com/SUPPORT/Pages/default.aspx

Try uninstalling them in Device Manager (as I can't find an installation for them elsewhere).

sygnus21 · 18 May 2020

@zbook,

Thanks for looking into this. In a funny kind of way I'm not surprised the GPU is being flagged as it's been acting weird since I got it last year. I probably should have RMA'd it, but for the most part it's been stable, especially while gaming.

@jdc1, disabling XMP is an option, but at this point in time I'll leave it in place and focus on the GPU.

As for Acronis, I'm actually running Acronis 2018 and have been using that version for about 2 years now. Not sure why ATI 2017 drivers are showing, especially since I did a clean install late last year and never ever had ATI 2017 installed.

Also, If you looked at my Windows version, you'd see all my "Windows updates" are installed. The BIOS is the latest, last BIOS for my board. I'm obsessive about driver and firmware updates so I have the latest of all of those unless something just recently released. That said I know there's a new driver release for my GPU so I'll check those later.

I'm also a system builder and routinely do quarterly cleaning/maintenance checks of my PC. I'm also pretty competent in my PC knowledge and build skills to know a lot of what you say; I just don't know how to read dump files (yet)

That said, since you don't know my knowledge level, you have to ask. With that, I appreciate all the tips.

All that said, I will take a hard look at the GPU.

Thanks guys

zbook · 18 May 2020

Multiple GPU drivers can be tried while testing computer stability / instability.

When testing older drivers turn off Windows updates of non-Microsoft drivers:
Enable or Disable Driver Updates in Windows Update in Windows 10

If multiple drivers do not achieve computer stability then check the motherboard for an alternative PCI-e slot and see if switching slots makes any difference.

If modifying the drivers and / or PCI-e slot achieve computer stability with normal temperatures then your set.

If there is still instability with multiple drivers then plan to swap test the GPU card.
The swap can be AMD or Nvidia.

When performing swaps see if you can perform two tests.
a) Test a new card in the problematic computer
b) Test the card from the problematic computer in another computer

jdc1 · 18 May 2020

Yes, IMO the GPU is the critical factor in the problems that you describe (because of the video stuff that I mentioned in my previous post).

Acronis release dates (such as Acronis 2018) aren't the same as the dates on the drivers.
As a matter of fact, in order to release ATI2018, they had to develop the drivers in 2017.
That's not an issue as far as I'm concerned - but getting updates for the program is a concern as there can be problems and unpatched vulnerabilities that are fixed in the updates. I'mnot saying to buy another version, just check at the Acronis website: Customer Service & Support - Acronis

The latest version of ATI 2018 is Build 15470 and it was released on February 5, 2019
https://download.acronis.com/AcronisTrueImage2018.exe

I apologize for asking about your Windows Update status. I have no easy way of telling which updates apply to your system - and which you have installed (this includes any optional updates which may apply to your system). AFAIK, Windows version number does not equate to which updates are installed. It is mainly a reference to the version of the kernel (ntoiskrnl.exe) that is being run on the system.

It is rare to find someone who regularly updates their system on the forums (and I did notice your status here at TenForums).
Most of what we see is OP's updating their system once the BSOD's hit - and most often we find that they've missed some critical updates in their haste to try and fix things. I, myself, don't regularly do updates - but I don't hesitate to "wipe and reinstall" if needed.

Reading dump files isn't that hard - at least in the beginning. It's the unusual commands in the debugger that throw most for a loop. The best command to use is !analyze -v The next most helpful commands are to list drivers, to see the stack, and to get information on drivers. Check for a PM from me with more about this...........

sygnus21 · 18 May 2020

zbook said:

Multiple GPU drivers can be tried while testing computer stability / instability.

Me: Yeah, I can always swap cards between my primary and backup system, but I'll probably be replacing this card if I have to go there since the card in the backup system was in this one and worked without issue.

When testing older drivers turn off Windows updates of non-Microsoft drivers:
Enable or Disable Driver Updates in Windows Update in Windows 10

Me: That's the first thing I do anytime I install Windows. It's also one of my recommendations

If multiple drivers do not achieve computer stability then check the motherboard for an alternative PCI-e slot and see if switching slots makes any difference.

Me: I'll give the latest drivers a try and go from there.

When performing swaps see if you can perform two tests.
a) Test a new card in the problematic computer
b) Test the card from the problematic computer in another computer

Me: see first reply in this post

Thanks gents. Now I know where to look and at least know it's not memory or CPU

Dude · 18 May 2020

I just want to add something to the great advice you already got.

You should disable XMP and set manually. Memory can be funny, and may be the real root cause here for TDR errors

sygnus21 · 18 May 2020

Dude said:

I just want to add something to the great advice you already got.

You should disable XMP and set manually. Memory can be funny, and may be the real root cause here for TDR errors

Yeah, I used to set manually back in my overclocking days, but with these newer boards there are so many options its just easier to set XMP and keep it moving.

Anyway, I'm 99.9 the issue is the GPU. As I said, earlier I should have RMA'd it last year. It runs and games well, but on rare occasions (and outside of gaming) I'll get some weird video issue. In the 14 months I've had the card I can count about 5 issues pointing to the GPU. There where no issues with my old card installed.

I guess I didn't want to RMA the card and wait for a replacement. Now I'm two months outside the RMA period