Display driver amdkmdag stopped responding and has successfully recovered.

I had exactly the same issue and for me the reason was one faulty modul of RAM. Problem occured both on RX 5700 XT and GTX 1050 with R5 1600X. After changing my RAM this problem was solved. So you can try swapping your RAM if you haven't done it yet. I'm now for a 3 weeks without driver timeout.

I have been struggling with this same problem, it started about a week ago. I have RX 570 card, AMD 8350 processor.  I was getting multiple crashes a day, no consistent pattern, could be gaming, could be waking up from sleep mode, could be just browsing files. Sometimes black screen, sometimes just everything froze.

I tried almost everything listed on this thread and nothing worked. 

I noticed that AMD released updated drivers on 2/1/2021. I did a complete uninstall of old drivers, installed new, still had crashes.

Last things I did before the problem was cleared up for at least 24 hours now:

• Put video card in other PCI slot
• Plugged different power cable into video card (one line with two connectors, I just switched to the other one)
• Moved CPU fan mounted to heat sink over a bit so that it wasn't touching RAM chips.

I don't know if or how any of these three may have solved the problem, but once I powered on after these changes, the frequent freezes stopped.

Just thinking about the last post and some of the other observations here - is it possible the cause of this problem is the card intermittently losing power? Even if there was a small voltage drop for just a microsecond - would that cause the clock to falter as the last poster observed?

With problems like these, sometimes it's a very small flaw in a wire or connector. This is what happened with my washing machine of all things - repair guy replaced several parts, still couldn't fix it. I got fed up, got my own multi-tester and the service manual, found intermittent connectivity in the main wiring harness, power to motor was getting cut off when the whole thing vibrated on spin cycle. I spliced the wires, and the washer runs 100% now. The repair guy was like - well I've never encountered that in blah blah blah years. Arg.

Anyhow, I'm thinking it could have been just a slight fault in either the pci connector, the power connector, or perhaps the cpu fan was pushing just a little bit against the ram chips.

If crashes resume, I will follow up with a picture of my PC being flung out the window.

I've ran furmark today twice, once for 30 minutes and then for around 7 minutes both times i had the driver timeout and this time rebooting didn't help. I checked Afterburner logs and found out that when the driver times out the memory clock goes from around 1700Mhz to 200 for a second, then jumps back up. Not sure if it artifacted when the driver crashed because i left the room with furmark running. I haven't had any issues with my pc for the last week but i wanted to make sure everything is fine, nd as it turns out, it isn't. I hope it's not a hardware issue with vram, i'm not too keen on struggling with the store i bought the gpu from and/or amd on the rma process. There was no temperature spike, VRAM usage spike, nothing out of ordinary. Just the memory clock drop.

I never tried my graphics card in another PCI slot on the motherboard, however I tried both HDMI and Display Port connections, on both my screens, and even 1 screen at a time.

As I mentioned in a previous post, the only fix I found was changing my USB peripherals setup.

I have keyboard/mouse plugged up top of my rear pannel, RGB mousepad in the following port a little lower, and my headsed plugged in the front panel.

I tried 2 different ports behind for my headset and/or RGB mousepad and I was getting crashes, with both my old Radeon RX 580 and my new GeForce RTX 3080 (so no, problem is not AMD exclusive)

However I haven't tried switching USB around since the latest Windows update release.