Frequent display driver crashes and MEMORY_MANAGEMENT BSOD

[cdawall]

Memory isn't on the QVL list for the motherboard. You can try a BIOS update and hope for the best, but this issue probably wont go away. No idea why OS choice is being brought up. Ryzen functions fine inside of 7.

 

[eidairaman1]

What he said. Not enough juice. Seems we're back to where we where yesterday.
Also increase VDDO_SOC voltage by 0.25 (SOC is integrated memory controller power settings). I can't see it listed in the book. It's in M.I.T. somewhere.

If a hsf is on cpu too tightly it can cause weird errors relating to memory due to board flex, it's very important to read board manual for what ram banks to use first if using just 2 modules too. A Bios update with latest AGESA can improve ram compatibility and stability too.

There is always a chance the ram is bad too

 

[TSR3]

to eidairaman1 - I think OPs GPU needs to downclock slightly after the memory is stable. As a test, to say, 480 default clocks.

Memory isn't on the QVL list for the motherboard. You can try a BIOS update and hope for the best, but this issue probably wont go away. No idea why OS choice is being brought up. Ryzen functions fine inside of 7.

Indeed. Page one of manual. Supports windows 7/64.

If a hsf is on cpu too tightly it can cause weird errors relating to memory due to board flex, it's very important to read board manual for what ram banks to use first if using just 2 modules too. A Bios update with latest AGESA can improve ram compatibility and stability too.

There is always a chance the ram is bad too

Also OP bought his 580 used. Might not be an issue but worth considering.

 

[INSTG8R]

@TSR3 quit double/triple posting you’ve already been told...

 

[TSR3]

@TSR3 quit double/triple posting you’ve already been told...

Out of here. I was just trying to help. Good luck Laser.

 

[LAserman367]

Bump ram voltages up, also get memtest 86

I did four passes before I made this thread, no problems. same with the shittier version that comes with windows itself.
Set it to 1.3V right now with XMP still enabled, if I understood my bios manual correctly it'll still be 1.2V. If that's the case I'll try to set the timings and speed manually as well
EDIT:: Yep still shows 1.2V for the RAM, going to manually set speed and timings now
EDIT2:: Wait no it shows as 1.308 in BIOS. I'm guessing CPU-Z readings might not be accurate? Gonna check if workers still die.
EDIT3:: Display Driver crashed instantly as I started the Blend test. After restarting, it ran for like 15 seconds then workers died one by one again.

What he said. Not enough juice. Seems we're back to where we where yesterday.
Also increase VDDO_SOC voltage by 0.25 (SOC is integrated memory controller power settings). I can't see it listed in the book. It's in M.I.T. somewhere.

if it's not in the book I imagine I can't do it. Could it be Dynamic VCORE SOC(DVID)?

If a hsf is on cpu too tightly it can cause weird errors relating to memory due to board flex, it's very important to read board manual for what ram banks to use first if using just 2 modules too. A Bios update with latest AGESA can improve ram compatibility and stability too.

There is always a chance the ram is bad too

I flashed the bios to latest version but I guess it was reset when I cleared my CMOS due to a previous issue0 Although it is still on a pretty recent version, since it shipped with that. Guess I'll do it again though.
As for the heatsink, I think it's on pretty lightly because I was very scared of it lmao. It's hard to tell because of those springs what's hard and what's not, and I don't have one of those screwdrivers that can measure pressure.

Memory isn't on the QVL list for the motherboard. You can try a BIOS update and hope for the best, but this issue probably wont go away. No idea why OS choice is being brought up. Ryzen functions fine inside of 7.

I picked this specific ram because it was listed together with the motherboard on logicalincrements. I thought the majority of incompatibility issues had been solved already and were only around for the first few months of ryzen. Though even if it's on the QVL wouldn't that simply imply it won't run at the rated speed?

 

[cdawall]

I picked this specific ram because it was listed together with the motherboard on logicalincrements. I thought the majority of incompatibility issues had been solved already and were only around for the first few months of ryzen. Though even if it's on the QVL wouldn't that simply imply it won't run at the rated speed?

You would assume that things were fixed, and it simply wouldn't work at rated speeds, but in practice they are just garbage still

 

[eidairaman1]

Send your gpu back and buy brand new

 

[LAserman367]

Send your gpu back and buy brand new

How certain are you it's due to the GPU? Amazon's "easy 30 day return" period ends in 9 days for the ram cpu and motherboard, so I probably won't have time to return any other parts in this time if I return the GPU and buy a new one to test. as
I contacted the seller the day about the problems and he hasn't seen the messages yet, so I contacted him on his number as well now.
Should I just slam the 580 in my old system and see if there's issues there? I'm not sure if I would even be able to force any issues due to a tremendous CPU bottleneck then though and I'm not sure if a lack of issues on that system would even indicate anything, as my r9 270x worked fine on that system too.

 

[eidairaman1]

Ok put the 270 in and test, if your issue goes away its the 580 gpu...

 

[LAserman367]

Ok put the 270 in and test, if your issue goes away its the 580 gpu...

if I put my R9 270x in, it doesn't boot. Hangs on third VGA light. But it works fine on my old system. see this older thread I made here
I bought this GPU because the 270x didn't boot at all
I have an even older zotac GTX470 which did boot, but I replaced it with the r9 270x in my old system because it also had frequent display driver crashes in the old system, which it still does and does as well on my new system.

 

[eidairaman1]

if I put my R9 270x in, it doesn't boot. Hangs on third VGA light. But it works fine on my old system.
I bought this GPU because the 270x didn't boot at all
I have an even older zotac GTX470 which did boot, but I replaced it with the r9 270x in my old system because it also had frequent display driver crashes in the old system, which it still does and does as well on my new system.

Told you before send the 580 back...

How about taking it to a shop then at this point.

 

[LAserman367]

Told you before send the 580 back...

How about taking it to a shop then at this point.

Not sure how that post makes you think I should return the 580 even more.
Also handing in pc to shop would probably mean I'm without a pc for like a month, miss out on my return periods and will probably end up with a few hundred euro bill at the end of it

Seller just got back to me and said he won't let me return it so there goes that option as well. I could try to sell it on to someone else though
But really, at this point I'm wondering if I send back just my ram to amazon or my ram cpu and motherboard

 

[eidairaman1]

You made things for yourself pretty difficult, do what you want.


/thread

 

[LAserman367]

Well an update, decided to fiddle around with the ram sticks. Pulled them both out and stuck one in a different colored slot and left the other out. Ran prime95 blend test and it ran 20 minutes on all workers, not a single one died and they all completed 1 or two tasks.
Shut down then plugged other stick in the paired slot, ran prime 95, workers started dying within a minute again
Pulled out both sticks, stuck second stick in the first stick's slot and left the first stick out this time (first is the one that tested fine on its own).
Started prime95 blend and instant deaths.

This should be a clear cut case of that stick being broken and an easy return no?

EDIT:: the single first stick seems to run fortnite just fine which crashed within 10-15 seconds every single time before.

 

[eidairaman1]

As long as you read your manual for different ram configurations, Id return the ram for replacement, after if it does the same ,replace the motherboard...

 

[Sempron Guy]

This happened to me twice as indicated on my screenshot above while stress testing my ram. I've used HCI Memtest, with memory load at around 94%. Same as what you've experience. Screen goes black for around 2 secs then saw that in the windows event viewer. I passed the HCI memtest run which I had running for around 16 hours w/o errors. Ran Passmark Memtest v7.8 after, passed without errors. Never happens to me though during normal usage. Only while stress testing the ram. Specs on my sig.