BSOD Issues

alpha0meqa

alpha0meqa

Member
Joined
Aug 13, 2023
Posts
7
I may have posted this wrong in another location. Apologies for that.

Hello all,


Let me preface this with saying, I don't know that this issue happens unless I'm using a sensor panel inside of my PC - it's acting as a "4th" monitor and just shows something from aida64.exe (my temps/fanspeed and etc). ---utilizing HDMI and 1 USB-A plug to my PC.---Plugged directly into my gpu.


I have some dumps and bsod messages as well that I will post. I will just randomly get BSOD though. It doesn't happen instantly or maybe even the same day. Could be days?


In addition to required files. I'm attaching an image/logs//dxdiag/ and a windbg run on the dump file, otherwise not sure how I can upload my dump file here. Lastly I have a picture of the BSOD from around 4 months ago - I could not get one from today.


I do have XMP enabled if that matters, and just bought an extra 32gb of ram(so 2 sticks to 4 sticks) exact same model and everything, but this issue happen before the ram increase too.



Questions required:
  • A brief description of your problem (but you can also include the steps you tried) - BSOD happens somewhat randomly, and not every minute/hour/day....
  • System Manufacturer? - self Built
  • Laptop or Desktop? - Desktop

  • OS ? (Windows 11, 10, 8.1, 8, 7, Vista) - Windows 11
  • x86 (32bit) or x64 (64bit)? 64bit

  • What was original installed OS on system? Windows 11
  • Is the OS an OEM version (came pre-installed on system) or full retail version (YOU purchased it from retailer)? I purchased it
  • Age of system? (hardware) 8 Months
  • Age of OS installation? 8 Months
  • Have you re-installed the OS? No
  • CPU 13700K
  • RAM (brand, EXACT model, what slots are you using?) - Added 32gb (2 more sticks recently) now using 4 slots, 16gb each dimm, and it's corsair vengeance 6000mhz DDR5 (XMP IS ENABLED in bios)
  • Video Card 6900XT
  • MotherBoard - (if NOT a laptop) - MSI Z690 EDGE
  • Power Supply - brand & wattage (if laptop, skip this one) - Corsair RM 1000e
  • Is driver verifier enabled or disabled? Disabled (I did run it on default settings once and had no issues found
  • What security software are you using? (Firewall, antivirus, antimalware, antispyware, and so forth) - Just Windows Defender
  • Are you using proxy, vpn, ipfilters or similar software? No
  • Are you using Disk Image tools? (like daemon tools, alcohol 52% or 120%, virtual CloneDrive, roxio software) Daemon tools once in awhile
  • Are you currently under/overclocking? Are there overclocking software installed on your system? ---Just XMP Enabled for ram

SPECCY:
http://speccy.piriform.com/results/fGXcNlquDtt6HWqxrmlzbXx
 

Attachments

I can't seem to figure out how to edit, but I ran the driver verifier, and nothing happened with that i guess? Also ran chkdsk and will attach those results.
 

Attachments

Hello, and welcome to the forum!

The first thing I have to admit is that I have no idea what you're describing here...
I'm using a sensor panel inside of my PC - it's acting as a "4th" monitor and just shows something from aida64.exe (my temps/fanspeed and etc). ---utilizing HDMI and 1 USB-A plug to my PC.---Plugged directly into my gpu.
Click to expand...
Could you be more descriptive and elaborate further on exactly what you're doing here?

The four dumps you uploaded suggest to me that this may well be a RAM issue. You have two 0xC0000005 exception codes (invalid memory access) and a non-cannmoical PTE entry, which may well be due to bad RAM. The outlier is a 0x119 which generally points at the graphics card or driver - that's why I'd like to know more about that 'sensor panel' setup.

You mentioned that this problem was happening before you added 32GB of RAM, so it's likely the original sticks in which the problem lies. Since you seem to have way more RAM than you need, the best way to check your RAM is to remove the two original sticks and just run on the two new ones. If it still BSODs then swap the RAM over and just run on the two original sticks (and with the new ones removed). If it BSODs on either pair of RAM sticks then RAM is not the issue and we'll need to look elsewhere.
 
I tested out all 4 dimms in the memtest86 and ran it over night with xmp on. It failed test 7 a few times. Trying now with just two sticks.

As far as sensor panel it's a little screen in my case that shows my temps. It works just like a normal monitor except it's like 1920x515 resolution.
VSDISPLAY 12.6" NV126B5M-N41 12.6inch 1920X515 LCD Screen Work with 2 HD-MI Mini LCD Controller VS-RTD2556HM-V1 https://a.co/d/glpbpMS

Also as a heads up. I bought one new kit of 64gb(2dimms)6400mhz and will try just using those two later on today when they get here.

The gpu issue... Idk if it's a problem or not. Idk how to test much with that
 
Any Memtest failure indicates bad RAM.

Out of interest, why do you need 64GB of RAM?
 
Funnily enough. I was hitting like 20-25 or so gb while gaming and watching streams. So I wanted a higher amount to not have to worry about it. Testing just two dimms now. I have done these tests before and I swear I think I only see failures when they are All together... I'm sure that doesn't make sense.
 
Ah. It's ALWAYS a mistake to look at the used RAM number and think that because it's high you must be running short. That's not the case. RAM is an expensive resource and to get the most benefit from that expensive resource you want it to be fully utilised. That's what the memory management component of Windows does - it maximises your use of RAM by keeping as much data in RAM as it possibly can, because the CPU can only directly access code and data that is in RAM. Yet it manages the use of RAM carefully so that when RAM is required for some new code or data it can be made immediately available.

The easiest way to see whether you are running out of RAM is to open the Task Manager, click the Performance tab, and then click the Memory icon. At the bottom there are some numbers, the ones you need are the Committed numbers. Processes run in virtual storage and whenever any process wants to use RAM it has to first allocate it by asking the Windows Memory Manager for permission to access the specified amount of virtual storage. When Windows allows the access it adds that value to the Committed value - so this is the total amount of virtual storage that the Memory Manager has guaranteed can be used across the system. When a process releases virtual stoarge the Memory Manager decrements the committed value by that amount of course.

Virtual storage exists in the real world in one of two places; RAM or the paging file.

In the Committed numbers in Task Manager, the right hand one (the bigger one) is the absolute maximum virtual storage that the Memory Manager can allocate at this moment - it's the sum of RAM and the size of the paging file (note that the paging file size can change). The left hand Committed number (the smaller one) is the current amount of virtual storage that the Memory Manager has guaranteed can be used. Clearly, the current committed value (the left hand number) must always be less than the absolute maximum (the right-hand number), because that one is an absolute limit. However, as long as the current committed value (the left hand number) is less than the size of installed RAM you are not running short of RAM. If the current committed value (the left hand number) is greater than the size of installed RAM consistently and for long periods, you may be running short of RAM and a RAM upgrade might be wise.

When you buy RAM you really want to buy a pack of matched RAM sticks. Your four RAM sticks must operate in synch, and that means they must be matched. Even though you've taken care to buy the same part number (which was very wise) they may not be completely matched, and at 6000MHz the slightest differences are going to become more noticeable. That may explain why those four RAM sticks don't play well together. You should be fine with the new 32GB pack because they will (should!) be a matched pair.
 
ubuysa said:
Ah. It's ALWAYS a mistake to look at the used RAM number and think that because it's high you must be running short. That's not the case. RAM is an expensive resource and to get the most benefit from that expensive resource you want it to be fully utilised. That's what the memory management component of Windows does - it maximises your use of RAM by keeping as much data in RAM as it possibly can, because the CPU can only directly access code and data that is in RAM. Yet it manages the use of RAM carefully so that when RAM is required for some new code or data it can be made immediately available.

The easiest way to see whether you are running out of RAM is to open the Task Manager, click the Performance tab, and then click the Memory icon. At the bottom there are some numbers, the ones you need are the Committed numbers. Processes run in virtual storage and whenever any process wants to use RAM it has to first allocate it by asking the Windows Memory Manager for permission to access the specified amount of virtual storage. When Windows allows the access it adds that value to the Committed value - so this is the total amount of virtual storage that the Memory Manager has guaranteed can be used across the system. When a process releases virtual stoarge the Memory Manager decrements the committed value by that amount of course.

Virtual storage exists in the real world in one of two places; RAM or the paging file.

In the Committed numbers in Task Manager, the right hand one (the bigger one) is the absolute maximum virtual storage that the Memory Manager can allocate at this moment - it's the sum of RAM and the size of the paging file (note that the paging file size can change). The left hand Committed number (the smaller one) is the current amount of virtual storage that the Memory Manager has guaranteed can be used. Clearly, the current committed value (the left hand number) must always be less than the absolute maximum (the right-hand number), because that one is an absolute limit. However, as long as the current committed value (the left hand number) is less than the size of installed RAM you are not running short of RAM. If the current committed value (the left hand number) is greater than the size of installed RAM consistently and for long periods, you may be running short of RAM and a RAM upgrade might be wise.

When you buy RAM you really want to buy a pack of matched RAM sticks. Your four RAM sticks must operate in synch, and that means they must be matched. Even though you've taken care to buy the same part number (which was very wise) they may not be completely matched, and at 6000MHz the slightest differences are going to become more noticeable. That may explain why those four RAM sticks don't play well together. You should be fine with the new 32GB pack because they will (should!) be a matched pair.
Click to expand...
Goodness that was a very good explanation. Thank you so much. And yes I tend to look at my % used for cpu and memory in task manager and in the actual task tab though so. That's my mistake for sure. I do use the performance tab occasionally and to confirm the xmp was working/ showing 6000mhz. But did not know all this information. I'll report back later and hope /pray this will fix my issue. I bought this hyte y60 case especially to use the sensor panel monitor inside and this build has just annoyed the hell out of me. Thank you!
 
Of course both memtest86 passed without issues with 2 sticks in at a time, on all 4 sticks :/ ---- do you think there could definitely be an issue with my GPU?
 
I thought you'd bought a new 32GB x 2 pack of RAM? Why are we still talking about 4 sticks?
 
I just wanted to test them before swapping. I swapped yesterday. I hope it was the ram somehow
Ram is a lot cheaper than gpu :)
 

Has Sysnative Forums helped you? Please consider donating to help us support the site!

Back
Top