Question Occassional BSOD during heavy task

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Aug 28, 2022
15
0
10
Around 2 years ago I build my first PC. I did not have an antistatic wriststrap to ground myself. Near the end of my build my monitor wouldn't turned on because the Ram wasn't fully plugged in so I fully plug it in while the PSU and the motherboard lights was on (I have Corsair Vengeance 2x8 3200 MHZ).

I had no problems with the PC for a while even though I did not ground and plug in Ram while PSU and motherboard lights was on, but after a while I would get BSOD occassionally for different error code each time and sometime my monitor would turn black for 2 second before turning back on. Or sometime the screen would turn into one color and I have to press Win + Ctrl + Shift + B to fix it. At the same time the monitor would get no signal if I accidently shake the HDMI cable too hard so sometimes I would be cautius of shaking the HDMI cable to not get bothered fixing the cord . I had a Ryzen 3 3200g with intergrated graphics so I thought the APU might did too much work so I brought gtx 1050 ti and the BSOD problem fix for a week and then it occur again but this time I only get PFN corrupt list and memory management error code for BSOD occasionaly after I install seprate GPU. Also when it goes to BSOD it go to 100% and it doesn't restart so I have to hold down the power button for 10 second to force shutdown.

I have Corsair Vengeance 2x8 3200 Mhz but when I set the speed to 3200Mhz the motherboard would boot on and off like 3 times before actually booting windows 10 on the monitor. One time I even accidently OC the speed to 3600 Mhz and the PC won't turn on at all so I had to clear CMOS.

Is there any ideas on which component I should replace because I would prefer not to replace my whole PC. Did not grounding myself fry my motherboard into damaging all the components connected to it or can some parts be saved.

Due to how messy are wires of my PSU, the wires also touch the metal part of the GPU and one wire is like resting on top of the pins of the GPU. I'm not sure if that's bad.

The PSU wires on the back of the case are also extending so far I can't close the back of the case so I'm not sure how bad is it to left the motherboard back pins in the air visibe. I may have touch the back of motherboard pins a couple times by accident.
 
Last edited by a moderator:

Colif

Win 11 Master
Moderator
Network drivers are LAN drivers, yes.


page file is a section of the ssd that the PC uses as virtual memory. All computers have more storage space than ram so for some actions, the PC uses the page file as extra space.

rebuilding it can move its location on the ssd and maybe avoid errors

Conversion of dumps

report - Click run as fiddle to see report

File: 100122-15515-01.dmp (Oct 1 2022 - 17:49:59)
BugCheck: [MEMORY_MANAGEMENT (1A)]
Probably caused by: memory_corruption (Process: WmiPrvSE.exe)
Uptime: 0 Day(s), 0 Hour(s), 02 Min(s), and 09 Sec(s)

File: 100122-14078-01.dmp (Oct 1 2022 - 17:53:54)
BugCheck: [MEMORY_MANAGEMENT (1A)]
Probably caused by: ntkrnlmp.exe (Process: NVIDIA Share.e)
Uptime: 0 Day(s), 0 Hour(s), 03 Min(s), and 25 Sec(s)

looking at your sysinfo from last crashes, 2 appear to have been caused by GPU drivers.

do you use GPU Tweak 2? this seems old
Apr 28 2017IOMap64.sys!!! ASUS 64-bit Input Output Map driver (GPU Tweaking driver)
It can also be part of AI Suite

Did you do this:
try running DDU in safe mode, remove GPU drivers, boot back into normal mode and either:
  1. install new drivers from Nvidia; or
  2. Run windows update and allow windows to find older Nvidia drivers that might be more stable
as you seem to have same drivers as before.
 
Aug 28, 2022
15
0
10
Network drivers are LAN drivers, yes.


page file is a section of the ssd that the PC uses as virtual memory. All computers have more storage space than ram so for some actions, the PC uses the page file as extra space.

rebuilding it can move its location on the ssd and maybe avoid errors

looking at your sysinfo from last crashes, 2 appear to have been caused by GPU drivers.

do you use GPU Tweak 2? this seems old
Apr 28 2017IOMap64.sys!!! ASUS 64-bit Input Output Map driver (GPU Tweaking driver)
It can also be part of AI Suite

Did you do this:
try running DDU in safe mode, remove GPU drivers, boot back into normal mode and either:
  1. install new drivers from Nvidia; or
  2. Run windows update and allow windows to find older Nvidia drivers that might be more stable
as you seem to have same drivers as before.
Ok so I ran Prime95 for not even an hour and I got these errors with 4 cores and blend as my setting.

View: https://imgur.com/a/xAqvwJs


I ran DriverEasy in the past, and got no erors, not sure how reliable that program is.

Since Prime95 give errors and Memtest didn't, am I supposed to replaced my CPU or Motherboard or RAM? I don't know if bad CPU can damage the motherboard or if bad motherboard can damage other components so if I buy a new CPU but turns out I have a damage motherboard then it damage the new CPU...

Also Prime95 test CPU but didn't I get BSOD error code like PAGE fault which means SSD or HHD or Motherboard?
 
Last edited:

Colif

Win 11 Master
Moderator
rounding error explains a lot

The way you fix rounding is by adding voltage to the Ram in bios. in small increments. Until it stops creating rounding errors.

I should have got you to check that before, it explains the bsod. Dumps aren't believable until its fixed.
am I supposed to replaced my CPU or Motherboard or RAM?
of those I would suggest RAM, but see if adding more power to it makes it happier.
your ram not running at 2133 until you added GPU is still odd. Wonder if replacing sticks, even though memtest shows them as fine, would fix the problem.

Memtest tests for errors in the cells of ram
Prime 95 tests CPU, RAM & Power.
Both test ram differently, can see different aspects of the problem, it seems its the power its getting that is cause. Adding a little more might be all it needs to run
 
Last edited:
Aug 28, 2022
15
0
10
Conversion of dumps


Not seen CCLeaner browser before, wonder if it deletes dump files like ccleaner does...

Okay, it seems that Asus update for LAN worked... sort of, its put even older drivers on now. Try running the link I gave you.
Might want to uninstall the LAN drivers in safe mode

on logon screen, click power button in bottom right
while holding left shift, click restart button
  1. this restarts PC in a blue menu
  2. choose troubleshoot
  3. choose advanced
  4. choose startup options
  5. click the restart button
  6. choose a safe mode (it doesn't matter which) by using number associated with it.
  7. Pc will restart and load safe mode
You might last a little longer there if problem is a driver.


on the off chance I am wrong, lets rebuild the page file
  1. go to settings/system/about
  2. in right hand side choose advanced system settings
  3. under Performance, click settings
  4. choose Advanced tab
  5. under virtual memory, choose change
  6. untick Automatically manage paging file size for all drives - this will let you select the areas that are currently greyed out
  7. select No paging File and choose set
  8. select ok and close out of all the windows
restart PC and follow all the steps until 8 and tick Automatically manage paging file size for all drives again to let windows manage it.

i would make the bootable Prime95 and run it off the USB. If the cause of BSOD is drivers, it won't happen when you boot off the USB. It can take a long time so I would run it overnight
I still get rounding errors in less than hour of running Prime95 and I changed my Ram voltage from 1.2V (Auto) to 1.25V to 1.3V to 1.4V. (Once one worker stopped then I close Prime95)

These are my motherboard settings.
View: https://imgur.com/a/nVBKibK

If I put voltage over 1.4V would it be bad? Or is my PSU need more watts.

Also I meant before getting a GPU I would get BSOD occassionally (no matter the Ram speed but it feels like 2133 mhz less BSOD then 3200 Mhz not sure, I put it on default so 2133 mhz to not mess with BIOS) and then when I get it more often I start to get a new GPU (with 2133 mhz Ram speed it didn’t BSOD for a month but with 3200Mhz it did BSOD). I suspected graphics problem due to sometimes no signal on monitor if I shake the caple by accident and sometimes screen glitch but ctrl win shift B fix the problem. That’s what happend.

I ran Prime95 again just now and it make 4 rounding errors in 0 mins pretty rare to be this fast.

Also ran Memtest86 again today, 4 errors in 2 minutes, 10 errors in 4 min. At pass 0, 2430 errors.
 
Last edited:

Colif

Win 11 Master
Moderator
I don't know if the cause is the ram or the Power Supply.
You can't keep throwing voltages at ram, its not good for it.
Its not working with memtest or Prime now.

Generally if I see errors in memtest I blame the ram.

Before i threw money at it blindly I would get PC tested by a repair store. Ask them to test PSU as its not the best model EVGA make.
 
Aug 28, 2022
15
0
10
I don't know if the cause is the ram or the Power Supply.
You can't keep throwing voltages at ram, its not good for it.
Its not working with memtest or Prime now.

Generally if I see errors in memtest I blame the ram.

Before i threw money at it blindly I would get PC tested by a repair store. Ask them to test PSU as its not the best model EVGA make.
I got new Ram, Motherboard, and CPU. Didn't want to get new Ram and break it if its bad motherboard. The GPU still works fine. No BSOD at all with Cinebench and Heaven Benchmark. Do I need to run Prime95 after Cinebench if I really want perfect stability? I don't want to have defective product without knowing til long after. I will test the new Ram with Memtest86 the next night, The old Ram show no signs of physical damage though and I wonder if I should get it replace by Corsair Limited Lifetime Warranty? I don't want to plug in bad ram to a brand new motherboard and break the other components. This time I try to be more cautious of grounding and eliminating as many possibility to damage components as possible so if a component is damage it would be more obvious. Should I ground myself when holding components (not installing) too?

PS: Taking the Heatsink out of the old CPU to get the CPU out was hard, if I apply so much force to the motherboard to take the heatsink out does it damage the motherboard? Just wondering for next time I change CPU on my new board feel like I'm damaging it.

This is what the old CPU looks like after taking it out View: https://imgur.com/a/0DyaXss
, from my view I see no bent pins so would it be labeled as defective? The missing pins in the corner and middle are meant to be there right?
 
Last edited: