How can I check from BSODs?

[Beinish]

Hello,
Lately, I've been having a lot of BSODs. After troubleshooting with a few experts, they decided it's probably a mobo > GPU > memory issue. Others said it may be the CPU as well.

I just swapped off the GPU and I still got a few game crashes but not BSODs.

I previously ran Prime95 with Small FFTs but there were no errors. I've been thinking maybe trying other tests, but after reading a bit online I understand that might be problematic? I saw many different opinions on which tests I should run and for how long.

I'd love your input on which tests to run to detect possibly faulty CPU/RAM.

My rig:
OS: Win 10 Pro (1909)
CPU: Ryzen 5 3600
Motherboard: ASUS TUF B450M Plus Gaming
Memory: Corsair Vengeance LPX 16GB (2x8GB) DDR4 DRAM 3200MHz C16
Graphics Card: EVGA RTX 2070 Super (Now using a GTX950 for testing).
PSU: Antec EA750G Pro 750W Gold
Hard Drives: Crucial P1 1TB 3D NAND NVMe PCIe M.2 SSD


Thanks

[rebirther]

What error code are you getting?

[Beinish]

Originally Posted by rebirther

What error code are you getting?

They vary, but here's a quote from a Microsoft MVP that helped me with the BSODs analysis after I sent him my dump and minidump files:

There are 3 different BugCheck in the 5 minidumps, 3 ea. 1E and 1 ea. 0A & 133, and that fact alone usually indicates a hardware problem.

The analysis of the memory.DMP file and the 5 minidumps do not indicate a software issue.

BugCheck 1E, {ffffffffc0000005, fffff80267858b53, 0, ffffe8f47a215890}
Probably caused by: hardware ( nt!MiUserFault+103 )

[rebirther]

If you have changed the GPU and using the same VGA driver without any issues you should get a new GPU. In some other cases it can also be the Mainboard BIOS or the power supplier.

[Beinish]

Originally Posted by rebirther

If you have changed the GPU and using the same VGA driver without any issues you should get a new GPU. In some other cases it can also be the Mainboard BIOS or the power supplier.

What do you mean the same VGA Driver? I changed drivers with the old GPU to test out if it's a driver error.

I haven't had a BSOD yet with the new GPU, just a game crash. It's a specific game that would crash on my previous and current GPU. The dev thinks it's some bug with Unity and my hardware. If the same game crashes with different GPUs, I doubt it's a GPU problem.

The problem is, it's hard to determine by BSODs because I didn't have one for a few days on the previous graphics card, and had not had one on the testing GPU either.

It might not be it at all but I will continue testing.

I don't know if it's indeed the mobo, but if not errors occur this week, I will send it to my local shop for them to test. I told them the mobo is probably faulty by they have to test it themselves to replace, which doesn't look good for me because I stress tested everything in my computer for hours.

[rebirther]

Originally Posted by Beinish

What do you mean the same VGA Driver? I changed drivers with the old GPU to test out if it's a driver error.

I haven't had a BSOD yet with the new GPU, just a game crash. It's a specific game that would crash on my previous and current GPU. The dev thinks it's some bug with Unity and my hardware. If the same game crashes with different GPUs, I doubt it's a GPU problem.

The problem is, it's hard to determine by BSODs because I didn't have one for a few days on the previous graphics card, and had not had one on the testing GPU either.

It might not be it at all but I will continue testing.

I don't know if it's indeed the mobo, but if not errors occur this week, I will send it to my local shop for them to test. I told them the mobo is probably faulty by they have to test it themselves to replace, which doesn't look good for me because I stress tested everything in my computer for hours.

You could try to install an older VGA driver for the RTX2070 and test it with Futuremark. What I have learned with win10 is in 99% of all cases its the driver.

[Beinish]

Originally Posted by rebirther

You could try to install an older VGA driver for the RTX2070 and test it with Futuremark. What I have learned with win10 is in 99% of all cases its the driver.

Done that with a few drivers. Each time uninstalling with DDU. Weird thing is that it started a few months back but the PC is barely 1 year old. Maybe something went bad after some time, maybe I needed to test more drivers. Frustrating how hard it is to determine the problem.

Anyway, I'm running more Memtest86 tests and was wondering if Prime95 would be able to stress test my CPU and RAM as well. That's the point of this post from the start.

[rebirther]

Originally Posted by Beinish

Done that with a few drivers. Each time uninstalling with DDU. Weird thing is that it started a few months back but the PC is barely 1 year old. Maybe something went bad after some time, maybe I needed to test more drivers. Frustrating how hard it is to determine the problem.

Anyway, I'm running more Memtest86 tests and was wondering if Prime95 would be able to stress test my CPU and RAM as well. That's the point of this post from the start.

If you have already tested older drivers then the GPU is damaged.

[retina]

Isolate and test each part separately.

Pare it down to a minimal setup.
One RAM stick. Test.
The other RAM stick. Test.
Both RAM sticks. Test.
With a GPU. Test.
With an HDD. Test.
etc.

If all those tests fail then it is something common like the PSU, the CPU, the mobo, the software, or too aggressive clocking/voltages.

[paulunderwood]

Adding to what retina advises, run memtest86+ and install a temperature monitoring software.

Also try running the memory at 3000Mhz.

[Beinish]

Originally Posted by retina

Isolate and test each part separately.

Pare it down to a minimal setup.
One RAM stick. Test.
The other RAM stick. Test.
Both RAM sticks. Test.
With a GPU. Test.
With an HDD. Test.
etc.

If all those tests fail then it is something common like the PSU, the CPU, the mobo, the software, or too aggressive clocking/voltages.

Originally Posted by paulunderwood

Adding to what retina advises, run memtest86+ and install a temperature monitoring software.

Also try running the memory at 3000Mhz.

I'll do the separate RAM test and as I mentioned, I got a another GPU to test as well.
I already ran countless RAM tests though. Memtest, Testmem5, HCI Design memtest, etc. I doubt the RAM is faulty to be honest. I suspect either the GPU, PSU, CPU or simply the mobo.

I already stress tested my previous GPU, ran Prime95, AIDA Extreme, OCCT for my PSU and GPU, and I believe I forgot to mention a few other stress testing tools.

I'm also not overclocking any component whatsoever. Everything is stock, Windows is formatted, latest drivers for BIOS and Chipset are installed.

I just wanted to use Prime95's other test (because I already ran "small FFTs) to see if it's still possibly the RAM/CPU.

I already contacted EVGA about a possible replacement, but the back and forth emails take time.

Also, I contacted my local hardware store that sold me the Mobo/CPU/PSU combo about a replacement but they asked me to bring the mobo+CPU to them for testing, but since I can't make it crash with any stress test, I'm assuming they'll send me home empty handed.