BSOD before Windows loads - Diagnosing / Dead RAM?

Swi1ch

Swi1ch

Swi1ch

Associate
Joined
19 Nov 2008
Posts
1,584
5900x
MSI B570S Tomahawk Max
Teamgroup ProDark 8PACK 32GB
3080TI
Corsair RM1000x

Started getting BSODs in August 2024. At the time I put the machine aside and I've only just gotten around to sorting it. On first power on it BSOD'd as soon as Windows loaded. I wiped the drives entirely and did a fresh install of W11. I then ran a bunch of OCCT tests, memtest86 passes, and some 3DMark for good measure - everything seemed to be fine. That was two weeks ago and it's been running great at everything (general work stuff, 3D modelling, gaming) up until this morning.

I had a BSOD an hour or so after power on. After the restart it instantly BSOD'd again after logging into Windows, and then after that it BSOD's before I can even get to the login screen. The BSOD codes seem to imply that the RAM may be at fault:
IRQL_NOT_LESS_OR_EQUAL (0XA) - ntoskrnl.exe
PAGE_FAULT_IN_NONPAGED_AREA (0X50)
KMODE_EXCEPTION_NOT_HANDLED (0X1C) - ACP.sys

I tried running with a single stick in slot 2 to see if one of them has died. Stick 1 would boot but BSODs before Windows login, stick 2 would not POST. Putting both sticks in would then boot but instant BSOD. On a second test run I can't get either stick in any slot to POST.

Is this sufficiently conclusive to call it the RAM at fault? I could do with being reasonably sure as I have lifetime warranty with OCUK on these sticks, but I'm a little concerned that with both sticks not working it may indicate something else at fault, and now I can't even boot the machine I can't run anymore tests.
 
Swi1ch said:
5900x
MSI B570S Tomahawk Max
Teamgroup ProDark 8PACK 32GB
3080TI
Corsair RM1000x

Started getting BSODs in August 2024. At the time I put the machine aside and I've only just gotten around to sorting it. On first power on it BSOD'd as soon as Windows loaded. I wiped the drives entirely and did a fresh install of W11. I then ran a bunch of OCCT tests, memtest86 passes, and some 3DMark for good measure - everything seemed to be fine. That was two weeks ago and it's been running great at everything (general work stuff, 3D modelling, gaming) up until this morning.

I had a BSOD an hour or so after power on. After the restart it instantly BSOD'd again after logging into Windows, and then after that it BSOD's before I can even get to the login screen. The BSOD codes seem to imply that the RAM may be at fault:
IRQL_NOT_LESS_OR_EQUAL (0XA) - ntoskrnl.exe
PAGE_FAULT_IN_NONPAGED_AREA (0X50)
KMODE_EXCEPTION_NOT_HANDLED (0X1C) - ACP.sys

I tried running with a single stick in slot 2 to see if one of them has died. Stick 1 would boot but BSODs before Windows login, stick 2 would not POST. Putting both sticks in would then boot but instant BSOD. On a second test run I can't get either stick in any slot to POST.

Is this sufficiently conclusive to call it the RAM at fault? I could do with being reasonably sure as I have lifetime warranty with OCUK on these sticks, but I'm a little concerned that with both sticks not working it may indicate something else at fault, and now I can't even boot the machine I can't run anymore tests.
Click to expand...
 
A few things I would try.

- Clear CMOS, remove the battery, disconnect from mains then hold power button down for 30 seconds to discharge any remaining power.
- Disable fast boot in Windows.

If this system has been sitting aside for over a year do you have another system you have been using? Could you swap the ram from another system? (if you have 1)

Have you tried the ram in slots 3/4 (if your motherboard has 4 x dimm slots) was unable to find the model posted as something else came out)

Only other thing is it could be the motherboard causing the issues or the CPU? Could try re seating the CPU and while at it checking pins etc.
 
I'll have a go at clearing the CMOS tomorrow. I can't disable fast boot in Windows as I can't get to Windows, or the BIOS.

I do have a system in use aside from this, but it's been ropey at best for a long time and I'm taking a don't break what ain't broke approach, so I'm loathe to touch it. I think I can borrow some DDR4 to test with instead.

I tried all of the dimm slots with no change.

I've manged to get access to a similar motherboard I can use for testing, I hope to give that a go over the weekend.

It's possible that it's the CPU but that seems extraordinariliy unlikely - that would make it the second time that this CPU has failed in this machine.
 
Can you tweak voltages in bios to get them working?

The RAM has a lifetime warranty, but it is now considered EOL (end of life) so you may not get replacements.

You would have to contact OCUK to verify this and to see what they can offer you.
 
freddie64 said:
You would have to contact OCUK to verify this and to see what they can offer you.
Click to expand...
They most likely will offer a refund if they no longer stock the hardware in question.

I just sent back a 4070Ti, they confirmed it was faulty and gave me a full refund.
 
Swi1ch said:
I'll have a go at clearing the CMOS tomorrow. I can't disable fast boot in Windows as I can't get to Windows, or the BIOS.

I do have a system in use aside from this, but it's been ropey at best for a long time and I'm taking a don't break what ain't broke approach, so I'm loathe to touch it. I think I can borrow some DDR4 to test with instead.

I tried all of the dimm slots with no change.

I've manged to get access to a similar motherboard I can use for testing, I hope to give that a go over the weekend.

It's possible that it's the CPU but that seems extraordinariliy unlikely - that would make it the second time that this CPU has failed in this machine.
Click to expand...
I would try the different ram first as its the easiest and quickest thing to test and then do the motherboard 2nd.

I have heard another user on here that had a 2nd CPU die in 1 system as well (can't remember the post now etc)

Edit - found the post, it was here. AMD 3600 first and then a 5700X3D. a few years between the 2 CPUs killing over but the 5700X3D was only 5 months old.

Do I keep frying CPUs?

Hi, a few years ago I had to return an AMD ryzen 5 3600 because it stopped working, now my 5700X3D has just done the same. The 3600 was tested by AMD and found to be faulty, however, could it be overclocking that has killed them both? I undervolted them both and they've both been watercooled...
forums.overclockers.co.uk forums.overclockers.co.uk
 
Last edited:
freddie64 said:
Can you tweak voltages in bios to get them working?
Click to expand...

I can't get to BIOS, the PC will power on, fans spin up etc but no display output at all. I've tried an alternate known-working GPU and no change.

ED209 said:
I have heard another user on here that had a 2nd CPU die in 1 system as well (can't remember the post now etc)
Click to expand...

That user was overclocking so at least there's some explanation for that - I've run completely stock on both of these. It's also a different motherboard than when the first CPU packed in.

I also would have thought if the CPU had died then I wouldn't have been able to get the PC back online again this time round, but maybe I'm just coping.
 
So I tried a different RAM kit, and then swapped in a different motherboard. It'll power up with fans spinning but no display output at all.

Looks like another dead CPU then. I'm kinda stunned that a non-overclocked Ryzen chip dies within a year of purchase, and then the replacement also dies within a year.
 
Swi1ch said:
So I tried a different RAM kit, and then swapped in a different motherboard. It'll power up with fans spinning but no display output at all.

Looks like another dead CPU then. I'm kinda stunned that a non-overclocked Ryzen chip dies within a year of purchase, and then the replacement also dies within a year.
Click to expand...
At least you have narrowed it down.

RMAA it and wait for a replacement
 
You must log in or register to reply here.
Back
Top Bottom