Question Please help me figure out this random black screen freeze/crash without bsod.

[luckygoku]

So since a long time I have been getting this random hard lock/black screen freeze sometimes on idle , sometimes on light load. Mostly on idle. Haven't seen on heavy load afaik.
Here are my specs
CPU : Ryzen 1700
GPU: Rx Vega 64 (reference)
MB : Asus x370 prime pro
PSU : Corsair rmx750W
RAM : Corsair vengeance 16x2 (3000mhz)
audio interface : focusrite 2i2 gen 1
dual monitor setup.
peripherals : midi keyboard , external hdd

So one thing to consider is that my gpu is slotted by this kind of forked cable from the psu :

https://imgur.com/a/RFD3JpM

View: https://imgur.com/a/RFD3JpM


I suspected the gpu the most so I put my ancient gpu from older system , radeon hd 6770 , for testing if the crashes stop.
And to my surprise they did for a week , or maybe I got lucky that it didn't happen , since those crashes are so random that the pc can go weeks without that crash and can also crash more than once in a single day.
I then went on to rma my gpu , after waiting a month I got a replacement yesterday.
All worked fine in gaming and stress tests , until just now , the same crash happened again while just idling.

This is what exactly happens when that crash happens , screens go black , if audio was playing prior to the crash it continues for a few seconds or gets stuck in a buzz , gpu power led goes off after a while. Keyboard loses power too , the rest of the system and the fans keep running until I reboot. system boots normally after reboot.

I looked around the forum and found out that there were people who had the first gen scarlett 2i2 audio interface also facing similar problems .... Does anyone here know of a similar problem that you solved ?
Is it the psu , or ram or mb or is the rma gpu also faulty ... idk what to do now. For ram I have tested using memtest with no errors.
Maybe it didn't crash with the old gpu because it uses way less power.

 

[Diddly]

I would guess power supply or motherboard. Have you got a different PSU you could try?

 

[drivinfast247]

Try running two separate power cables from the PSU to GPU.

 

[TreborG2]

Completely uninstall AMD's drivers (if it asks if you want to delete.. YES!)
Restart windows 10
Let Windows download its video drivers (should still show your card's name)
ONLY use Microsoft's drivers for a while and see if the problem is gone.

I built my system in Feb/Mar 2018, 32G ram, Radon RX580 w/8Gram, amd 1700x .. ... I couldn't last a day with AMD's drivers in stalled.

usually I get a black screen .. if I catch it quickly .. I can do CTRL+ALT+DEL get the security screen up and select Task Manager .. if I can do that .. I can usually keep it up a bit longer.

Since completely uninstalling AMD's driver, and letting windows do its thing .. I may not have the highest frame rates, and I don't get the overlay and their replay bar, but I could go stable for days without rando video GPF's forcing me to reboot.

See if that helps ..

 

[luckygoku]

I would guess power supply or motherboard. Have you got a different PSU you could try?

I have another one from my older system. But its only 550W (seasonic) which I don't think is enough for Vega
I am now running two separate pcie power cables from the psu but the crash still happens. Temps got better by 5C for the gpu with the separate wires though.
Is it possible that mem86 fails to catch errors in some cases ?

 

[luckygoku]

Completely uninstall AMD's drivers (if it asks if you want to delete.. YES!)
Restart windows 10
Let Windows download its video drivers (should still show your card's name)
ONLY use Microsoft's drivers for a while and see if the problem is gone.

I built my system in Feb/Mar 2018, 32G ram, Radon RX580 w/8Gram, amd 1700x .. ... I couldn't last a day with AMD's drivers in stalled.

usually I get a black screen .. if I catch it quickly .. I can do CTRL+ALT+DEL get the security screen up and select Task Manager .. if I can do that .. I can usually keep it up a bit longer.

Since completely uninstalling AMD's driver, and letting windows do its thing .. I may not have the highest frame rates, and I don't get the overlay and their replay bar, but I could go stable for days without rando video GPF's forcing me to reboot.

See if that helps ..

I suspected AMD drivers but this seems like a hardware issue because there are no dump files created for these crashes/freezes.
Although it is true that this issue gets amplified on some driver releases. I always do DDU though before updating and also now the factory reset from amd in the new driver.

 

[luckygoku]

Try running two separate power cables from the PSU to GPU.

I did that , it helped with the temps a bit but the crash still happened ! Completely out of ideas now.
Maybe I should start with RMA'ng RAM first , then the PSU. I am not sure if the RMA gpu is also faulty or not.