AB350M Pro4 - Random Freezing/Lock Ups

Seeing as we're well past NewEgg's return period, I figure the only chance I have of recouping is getting this thing replaced and then reselling it. It'll be interesting to see how long it takes to get anything done.

Over 2 weeks now with zero response from support. Guess it's time to start leaving reviews...*those* they seem to respond to. (grumble)

Anyway, the problem is/was definitely rooted in the board.
I took a gamble on a cheap refurb MSI B350 Bazooka, and it's been running perfectly for the past couple days. Only time will tell, but it looks good for now.

During disassembly, I noticed that the VRM heatsink above the CPU was not fully secured, and rocked freely. (The pins were locked in, but the thermal pad obviously had no hold there, like the other one did.) Possibly part of the problem, though I was unable to observe any actual readings to confirm this.

Either way, it's time for them to honor their warranty.

Bringing this back over here, because I never should have hijacked Mati98's thread, and don't want my continued troubleshooting to get things even more convoluted.

@ xhue:
I took your suggestion to contact support. Been a few days, still waiting on a response.

In the meantime, I keep looking for possible answers.
I have some HWinfo logs taken before/during a crash... I thought they'd offer some hard evidence, but honestly, I'd don't know quite what I'm looking at or looking for. I'd happily post them if someone wants to take a gander.

I've had two unique crashes in the past day or so that might offer some further insight based on behavior...

The first unique crash was actually a "near-crash":
While playing some Neverwinter and running HWInfo waiting for a crash, the sound started crackling and the inputs and screen started to freeze per a normal crash -- but then, I noticed my keyboard (a backlit LED KB), completely shut off. It flickered on and off a few times, and the video stuttered along....and then everything seemed to recover.
I tabbed out of the game, stopped the current HWInfo log and restarted a new log. Went back into the game for a couple minutes to wrap up, then exited the game. Within 0-3 seconds of exiting, the system froze up at desktop.

This particular crash is also the first one to generate any kind of obviously related information in Windows Event Viewer -- not flagged as an Error or Warning, but just Information from Windows Error Reporting. A Live Kernel Event, a crash of the usb extensible host controller, which explained the keyboard behavior.

After this, I did some further testing. Realized I hadn't tried removing the NVME drive and using a standard SATA. So I did that. Still crashes, but....

The second unique crash was actually very much the same as the others.
In this instance, I was watching using Plex Media Player. Like "normal", inputs froze, video froze, no HDD LED, but audio continues normally.
But then, for whatever reason this time, I noticed that prior/during this crash, the activity/wait icon (the circle that spins next to your mouse cursor), was still spinning. So, the screen was frozen....yet not actually frozen? It breaks my mind.
No noteworthy entries in the event viewer this time.

I feel like this must be some sort of power fluctuation at the motherboard level, but I can't seem to find/prove it.

Thomas' RAM was also chosen from the QVL, yet he had a bad stick.

6 months strong without any (known) jolts or hard vibrations, plus a sudden instability mosr likely means some part is transcending to the Eternal Silicon Valley.

Such lockups that you have are rarely SW related, although Win 10 updates tend to break almost everything.

You may also try running only 1 stick of RAM for the sake of testing. Recording temp and volt values with HWiNFO is also an idea, just to check for abnormalities before the lockups occur. Maybe Win 10 error logs and events will shed some more light too?

My bet is incompatuble or loosely seated RAM.

Check a thread from thomas1986. Guy had same mobo and RAM as you do. Even his CPU is almost the same.

Try the steps described in the thread I mentioned and let us know how it went.