Application Crash dwm.exe ntdll.dll - Win 10

[DrunkBlood]

I can rule out RAM tested with a different set of sticks nothing has changed

[phillpower2]

It could be the memory slots themselves.

Entirely your choice what you test and I am more than happy to wait for MSI to get back to you.

[DrunkBlood]

So I tested my old ram alternating the ram itself and the mem slots. With no memtest86 error with atleast 4 passes I still tried new ram with no change to my problems. I do not know wether this rules out my mem slots.

[phillpower2]

What have MSI said?

Makes the RAM less likely to be the cause but not 100%, for conclusive results each stick has to be tested on its own for 4 or 5 passes and in each slot, we will look elsewhere for now, the GPU & checking the SSD..

In case you are not already aware, memory problems includes VRAM which is what your GPU uses and you having problems running Superposition could be a clue as to the cause if the issue, check the VRAM by running OCCT which is free for personal use, download from here.

So we can check the SSD;

Download and run CrystalDisk info standard edition from here

Run the program, grab any screenshots and attach to your next reply for us.

To capture and post a screenshot;

Click on the ALT key + PRT SCR key..its on the top row..right hand side..now click on start…all programs…accessories…paint…left click in the white area …press CTRL + V…click on file…click on save…save it to your desktop…name it something related to the screen your capturing… BE SURE TO SAVE IT AS A .JPG …otherwise it may be to big to upload… after typing in any response you have… click on Upload a File to add the screenshot.

Screenshot instructions are provided to assist those that may read this topic but are not yet aware of the “how to”.

[DrunkBlood]

For MSI I just got an automated response saying that they have alot of stuff to deal with as of right now and it will take a while.

I did CrystalDisk, OCCT CPU an hour and tried the OCCT GPU but could not get it to run for more than 2 min till the crash. But the CPU test was stable for the duration even showing ~ 37% GPU load.

SSD:
[ATTACH type=“full”]6126[/ATTACH]
[ATTACH type=“full”]6127[/ATTACH]

Voltage screen after some tries of GPU test:
[ATTACH type=“full”]6128[/ATTACH]
CPU is just attached

[phillpower2]

GPU failure again

Final test, your GPU manufacturers own testing program can be downloaded from here

[DrunkBlood]

This morning I did some tests with the tool from MSI and geek3d. Here are the results all tested with 1290x720 and standard settings
[ul]
[li]test info → time till crash[/li][li]1080 benchmark → 1min[/li]li msi-01 → <1s[/li]li msi-01 → at 60% load time[/li]li msi-01-> 20s[/li]li msi-02 medium++ → 15s[/li]li msi-02 medium++ → 17s[/li]li msi-02 medium++ → BSOD - Critical_Process_Died[/li]li furmark-msi → 2:30 low VRAM 512MB/8GB[/li]li furmark-msi → 3:14 low VRAM 512MB/8GB[/li]li furmark-donut → 2:40[/li]li furmark-donut-3200VRAM → 10s[/li]li furmark-donut-5300VRAM → 9s[/li]li TessMarkx32 → 1:10 low VRAM 232MB/8GB[/li]li Phong-Donut → 45:00 no crash low VRAM 256MB/8GB (screen attached, nvm 2.4MB is to large and ziping reduces file size by lower than 1%, imgur upload)[/li][li]I think (GL) means OpenGL and (VK) means Vulkan[/li][/ul]
[MEDIA=imgur]S28Kl7D[/MEDIA]

[phillpower2]

Right after looking at the above test results and then going back to the other previous GPU test results I decided to read your thread from the beginning so everything was fresh in ones mind.

I did recall mentioning to you way back that I suspected that there was two issues going on and what has been going on with your computer and what has been done already confirms that there was, the low storage space on the C: drive was one problem (software) that issue was initially resolved by you freeing up some space on the drive and then ultimately by you clean installing Windows, task one completed.

So that you have the post fresh in your mind;

Looked at the two latest crash dmps and they were both caused by different things, the latest one is a generic crash dmp that does not point towards a driver issue but more a hardware problem, the second crash dmp which is of no real use was caused by a conflict between Windows Defender and the Realtek network adaptor driver, it hasn`t happened since so that issue looks to have been resolved and possibly by one of the updates that you installed after the crash on January 10th 2020.

Now to cut to the chase, we have been testing the GPU as that is what keeps getting flagged up and after re reading the thread again I may have found the cause.

So that I don`t potentially end up with egg on my face can you see if I have the correct GPU MSI RX 5700 8G

[DrunkBlood]

Just took a photo of the box the GPU came with. Link to msi
[MEDIA=imgur]yeXIjlR[/MEDIA]

[phillpower2]

Confirms what I suspected after going through the recent GPU test results.

Did you not do any research before pairing such a GPU with only a 500W PSU, being a good brand of PSU using it will hopefully not have damaged the GPU any but being some 250W short of what MSI recommend you best prepare yourself for the worst

[DrunkBlood]

I summed all of the listed watts and came with about (65CPU+180GPU+Rest)300W where then I thought a buffer of 150W is enough. I also read online that most buy to high for their PSU
[ATTACH type=“full” alt=“1586880986873.png”]6154[/ATTACH]

[phillpower2]

Sorry but you should always be guided by what the GPU manufacturer recommends and as a minimum, many do overstate the power requirement to cover themselves but at the end of the day they are the ones you need to try and get an RMA from should something go wrong and if you have not used at least what they recommend to power the GPU then I`m afraid you do not have any warranty claim if the card breaks, their techs can tell as well btw.

The PSU requirements are widely available as well, one such example here

[DrunkBlood]

Then it was my bad only looking at PSU manufacturer and AMD page not msi

[DrunkBlood]

If it is final with no warranty for my GPU I would like to keep my power and downgrade to suitable GPU if you have some suggestions this would be helpful.

[phillpower2]

Don`t be too hard on yourself, no one is perfect.

Just a btw, I did notice two things in your Dark Power Pro screenshot.

The GPU is not the same one that you have and only the 650W PSU would be enough for a stock RX 5700 as they require 600W so following the PSU brands information would get you coming up short in any event.

If me, I would get an adequately powered PSU and keep the present GPU