BSOD after PC put to sleep connected to XMP

[Chris Matters]

Good evening Gentleman,

 

Let me point out at first I do have workaround so this is low priority issue.

....and yet it baffles me as I would like to understand why does it happen.....

 

Here it is:

Whenever I put my PC to sleep and then reboot it, it will sooner or later BSOD.

Moreover, it will keep doing these BSODs even after full reboot.

 

Here is interesting part.

1. If I remove one out of 2 memory sticks I have from mobo, system will run stable (if I re-add 2nd stick it is back to BSODs again)

2. If I disable XMP and run it without it (2 memory sticks used), it will run stable.

3. If I disable XMP profile and then re-enable it back again system will run stable.

 

As such, as a workaround I disabled automatic sleep on my pc.

Recently, I forgot about above and put PC to sleep manually so it could kindly remind me the issue is still there.

 

I honestly don't mind it as much but I would like to know root cause behind such strange behavior.

 

It looks to me like the fault lies with memory.

I ran memtest during issue and it said something was wrong and yet it was fine when i stopped sleeping my PC.

 

As for my RIG it is:

AMD X470 AORUS

Trident Z RGB DDR4-3200MHz CL16-18-18-38 1.35V 16GB (2x8GB)

2nd Gen Ryzen™ 7 2700X

[WaggishOhio383]

What are the BSODs you're getting?

[Chris Matters]

2 BSODs before I applied workaround.

If memory server me right it was various BSODs previously as well.

 

DRIVER_IRQL_NOT_LESS_OR_EQUAL (d1)
An attempt was made to access a pageable (or completely invalid) address at an
interrupt request level (IRQL) that is too high.  This is usually
caused by drivers using improper addresses.
If kernel debugger is available get stack backtrace.
Arguments:
Arg1: fffff802ab516678, memory referenced
Arg2: 0000000000000002, IRQL
Arg3: 0000000000000008, value 0 = read operation, 1 = write operation
Arg4: fffff802ab516678, address which referenced memory

SYSTEM_THREAD_EXCEPTION_NOT_HANDLED_M (1000007e)
This is a very common bugcheck.  Usually the exception address pinpoints
the driver/function that caused the problem.  Always note this address
as well as the link date of the driver/image that contains this address.
Some common problems are exception code 0x80000003.  This means a hard
coded breakpoint or assertion was hit, but this system was booted
/NODEBUG.  This is not supposed to happen as developers should never have
hardcoded breakpoints in retail code, but ...
If this happens, make sure a debugger gets connected, and the
system is booted /DEBUG.  This will let us see why this breakpoint is
happening.
Arguments:
Arg1: ffffffffc0000005, The exception code that was not handled
Arg2: fffff8066904b174, The address that the exception occurred at
Arg3: ffffaf027541b348, Exception Record Address
Arg4: ffffaf027541ab90, Context Record Address
 

[WaggishOhio383]

I also had the driver_irql one on my PC, and it turned out that my bios setting weren't providing enough voltage for my ram (1.2 v when it needed 1.35 v). Maybe check the recommended voltage for your RAM and make sure it's getting enough power? Also, have you tried running the windows memory diagnostic tool?

[Chris Matters]

I am using XMP profile and it sets Voltage to correct 1.35 v

I used memtest and it found issues but ONLY after I put PC to sleep.

When workaround was applied (reset XMP profile) memtest was passed fine.

 

To add a bit more information....

When I cleared CMOS and then loaded it back from drive, issue was still there.

 

To "fix it"

I have to go to BIOS disable XMP profile

Boot into windows, then shutdown

Go back to BIOS re-enable XMP profile

And boot into windows yet again.

[WaggishOhio383]

Odd... Maybe something is messing up your BIOS when your computer is in sleep mode? It couldn't hurt to try disabling fast startup if you haven't already as that can cause issues on some computers

 

https://help.uaudio.com/hc/en-us/articles/213195423-How-To-Disable-Fast-Startup-in-Windows-10

[Chris Matters]

I already have tried that one as well.

I actually left it disabled as its easier getting into BIOS without it.

[WaggishOhio383]

1 minute ago, Chris Matters said:

I already have tried that one as well.

I actually left it disabled as its easier getting into BIOS without it.

Well then I don't really know... I've exhausted all the methods I've used to fix similar problems before. This issue requires someone with more computer expertise than me. Sorry I can't help any further.

[Chris Matters]

Thank you for trying. I am not an expert either but I do consider myself rather experienced in troubleshooting.

As I mentioned, don't worry too much. As long as I'm not putting my PC to sleep, everything is fine.

 

More than solution, I am more curious why?

Maybe it is some kind of bug in bios (tried upgrading it) or RAMs SPD, god knows.

 

It is really odd... Remove one stick and system is stable even after sleep.

Maybe its memory controller?

[Chris Matters]

TLDR; Solution was to under-volt my RAM VDD to 1.31 V and VTT to 0,655 V

 

 

Good evening once again Gentleman,

 

I was neglecting this issue for 2 years.

I had workaround in place (do not put my PC into sleep mode) and didn't really care for solution...

 

Until recently, where I found same ram model for sale, half its price.

I decided "Why not to upgrade?", well aware of issues I had before.

 

And as I was afraid, the issue reoccurred.

Moreover, my workaround no longer did its magic...

I decided it is time to finally crack this on and find root cause.

 

And so I started, basics first.

I disabled XMP and downloaded newest BIOS from Gigabyte.

But that would be too easy now, wouldn't it?

 

So after confirming issue was still there, I started to wonder...

"How far can I push the speed before it becomes unstable?"

• 2400 MHz - stable,
• 2800 MHz - stable,
• 3000 MHz - stable,
• 3066 MHz - unstable,
• 3133 MHz - FAIL

At this point, I was already satisfied. 3 GHz compared to 3.2 GHz ain't that bad...

Probably wouldn't even notice the difference.

 

But I've decided to keep going. For science! For art! For fun!

 

And so my journey continued...

Suggested on many forums, I've decided to fiddle with voltage parameters.

I bumped it little by little up to 1.38V, my PC kept on crashing regardless.

I was considering going up to 1.4V, (and not a step further) but then I noticed something peculiar...

 

When I was checking on sensors in BIOS, I noticed VDD was jumping quite a lot.

This was interesting! At first I thought the reason for instability could be voltage jumping too much.

But I also noticed another thing. It never went to specified value. It was always above it...

 

This gave me a glimmer of hope. Last chance to redeem my G.Skill memory sticks!

So I set RAM speed back to 3066 MHz, this time together with voltage set down to 1.34V

And it was stable!

 

You probably know the rest of it (well duh, I added TLDR at the top).

I continued my tests. Whenever I received any signs of instability I degreased voltage by 1 point.

Baby steps, like last time. And here are the results:

3066 MHz. @ 1.34V - stable

3133 MHz. @ 1.33V - stable

3200 MHz. @ 1.32V - unstable

3200 MHz. @ 1.31V - STABLE AT RATED SPEED!

 

This concludes my story. I achieved desired speed and my system is stable.

I even verified putting PC to sleep will not break anything. (It looks stable so far) I am running tests as I am writing this.

At this point, after everything is finally over I suspect problem was with VRMs on X470 AORUS mobo (not cool Gigabyte)...

 

I was reluctant to write in this old post I created but I wanted people to know.

Sometimes increasing voltage might not be correct answer.

Sometimes you need to take things slow...

AND CHECK THE DAMN SENSORS!

 

PS. I also I written this because I was bored during my stability run....