Question Help needed! New build crashes. Over thousands of errors in OCCT CPU test.

[blablaalb]

So few months ago I built a new system. Since I was a bit short on money I decided not to buy a new GPU and stick with the old one (GT 620). The built came out extremely unstable, I had a lot of different BSOD. Since every component in the system was brand new I blamed the GPU for all the crashes. I bought new a new GPU few days ago, installed it, but the system still crashes. Today I ran OCCT test on RAM and it crashed. Then I ran CPU test and within few minutes it reported for over 3k errors and then suddenly my PC BSODed. I'm looking for a help. I don't know what can be the cause for all the BSODs. I updated all drivers on my PC with Driver Easy. I update BIOS to the latest version, went to the UEFI settings enabled the XMP 2.0 profile, but the system still crashes.

I have ran out of guesses, I don't know what may cause all of the BSODs. Maybe I need to configure CPU settings in UEFI?
Most of the errors are MEMORY_MANAGMENT, PAGE_FAULT_IN_NONPAGES_AREA and IRQL_NOT_LESS_OR_EQUAL.
Here is the content of one the dump files analyzed with WinDbg:

Spoiler: Minidump

Microsoft (R) Windows Debugger Version 10.0.20153.1000 AMD64
Copyright (c) Microsoft Corporation. All rights reserved.


Loading Dump File [C:\Windows\Minidump\022621-1268281-01.dmp]
Mini Kernel Dump File: Only registers and stack trace are available


Path validation summary *
Response Time (ms) Location
Deferred srv*
Symbol search path is: srv*
Executable search path is:
Windows 10 Kernel Version 19041 MP (12 procs) Free x64
Product: WinNt, suite: TerminalServer SingleUserTS
Edition build lab: 19041.1.amd64fre.vb_release.191206-1406
Machine Name:
Kernel base = 0xfffff8051ba00000 PsLoadedModuleList = 0xfffff8051c62a390
Debug session time: Fri Feb 26 17:07:15.284 2021 (UTC + 4:00)
System Uptime: 0 days 0:38:29.974
Loading Kernel Symbols
...............................................................
................................................................
............................................................
Loading User Symbols
Loading unloaded module list
.......
For analysis of this file, run !analyze -v
nt!KeBugCheckEx:
fffff8051bdf5a80 48894c2408 mov qword ptr [rsp+8],rcx ss:0018:ffffe70fa81481e0=0000000000000050
10: kd> !analyze -v
***

• *
• Bugcheck Analysis *
• *

***

PAGE_FAULT_IN_NONPAGED_AREA (50)
Invalid system memory was referenced. This cannot be protected by try-except.
Typically the address is just plain bad or it is pointing at freed memory.
Arguments:
Arg1: ffff8b9aedbf8050, memory referenced.
Arg2: 0000000000000002, value 0 = read operation, 1 = write operation.
Arg3: fffff8051bc611d6, If non-zero, the instruction address which referenced the bad memory
address.
Arg4: 0000000000000002, (reserved)

Debugging Details:
------------------


Could not read faulting driver name

KEY_VALUES_STRING: 1

Key : Analysis.CPU.mSec
Value: 4061

Key : Analysis.DebugAnalysisProvider.CPP
Value: Create: 8007007e on DESKTOP-3MF977B

Key : Analysis.DebugData
Value: CreateObject

Key : Analysis.DebugModel
Value: CreateObject

Key : Analysis.Elapsed.mSec
Value: 18195

Key : Analysis.Memory.CommitPeak.Mb
Value: 77

Key : Analysis.System
Value: CreateObject

Key : WER.OS.Branch
Value: vb_release

Key : WER.OS.Timestamp
Value: 2019-12-06T14:06:00Z

Key : WER.OS.Version
Value: 10.0.19041.1


ADDITIONAL_XML: 1

OS_BUILD_LAYERS: 1

BUGCHECK_CODE: 50

BUGCHECK_P1: ffff8b9aedbf8050

BUGCHECK_P2: 2

BUGCHECK_P3: fffff8051bc611d6

BUGCHECK_P4: 2

READ_ADDRESS: fffff8051c6fb390: Unable to get MiVisibleState
Unable to get NonPagedPoolStart
Unable to get NonPagedPoolEnd
Unable to get PagedPoolStart
Unable to get PagedPoolEnd
fffff8051c60f330: Unable to get Flags value from nt!KdVersionBlock
fffff8051c60f330: Unable to get Flags value from nt!KdVersionBlock
unable to get nt!MmSpecialPagesInUse
ffff8b9aedbf8050

MM_INTERNAL_CODE: 2

BLACKBOXBSD: 1 (!blackboxbsd)


BLACKBOXNTFS: 1 (!blackboxntfs)


BLACKBOXWINLOGON: 1

CUSTOMER_CRASH_COUNT: 1

PROCESS_NAME: firefox.exe

TRAP_FRAME: ffffe70fa8148480 -- (.trap 0xffffe70fa8148480)
NOTE: The trap frame does not contain all registers.
Some register values may be zeroed or incorrect.
rax=0000000000062970 rbx=0000000000000000 rcx=ffff8b9aedbf8080
rdx=00000000746c6644 rsi=0000000000000000 rdi=0000000000000000
rip=fffff8051bc611d6 rsp=ffffe70fa8148610 rbp=ffffe70fa8148709
r8=0000000000000003 r9=00000000ffffffff r10=fffff8051bc48710
r11=ffffe70fa8148768 r12=0000000000000000 r13=0000000000000000
r14=0000000000000000 r15=0000000000000000
iopl=0 nv up ei pl zr na po nc
nt!ObfDereferenceObjectWithTag+0x26:
fffff8051bc611d6 f0480fc15ed0 lock xadd qword ptr [rsi-30h],rbx ds:ffffffffffffffd0=????????????????
Resetting default scope

STACK_TEXT:
ffffe70fa81481d8 fffff8051be1ebbb : 0000000000000050 ffff8b9aedbf8050 0000000000000002 ffffe70fa8148480 : nt!KeBugCheckEx
ffffe70fa81481e0 fffff8051bc0c960 : ffffb600e8800180 0000000000000002 ffffe70fa8148500 0000000000000000 : nt!MiSystemFault+0x1f43ab
ffffe70fa81482e0 fffff8051be03c5e : 0000000000000000 0000000000000000 ffffb600e88af180 0000000000000cbe : nt!MmAccessFault+0x400
ffffe70fa8148480 fffff8051bc611d6 : 0000000000000001 000000005246704e ffffe70fa81487a0 0000000000000000 : nt!KiPageFault+0x35e
ffffe70fa8148610 fffff8051bc6119e : ffff8b8af2204850 0000000000000000 000000005246704e 0000000000000002 : nt!ObfDereferenceObjectWithTag+0x26
ffffe70fa8148650 fffff8051bc48e52 : ffff8b8ae59d9760 0000000000000000 0000000000000040 fffff8052c58dc9a : nt!HalPutDmaAdapter+0xe
ffffe70fa8148680 fffff8051bc48727 : 0000000000000001 0000000000000002 ffff8b8af1339b50 ffff8b8af006c388 : nt!IopfCompleteRequest+0x712
ffffe70fa8148770 fffff8052c58d834 : fffff80500000000 00000172110cfb40 ffff8b8af006c270 ffff8b8aeedac080 : nt!IofCompleteRequest+0x17
ffffe70fa81487a0 fffff8051bc52f55 : ffff8b8ae59d9760 ffff8b8ae59d9760 0000000000000001 ffff8b8af006c270 : Npfs+0xd834
ffffe70fa8148810 fffff8051afe4906 : 0000000000000000 0000000000000000 0000000000000000 fffff8051afe4906 : nt!IofCallDriver+0x55
ffffe70fa8148850 fffff8051bc52f55 : ffff8b8af006c270 0000000000000001 ffff8b8af1339b50 fffff80500000000 : FLTMGR!FltpDispatch+0xd6
ffffe70fa81488b0 fffff8051bffd878 : ffff8b8af006c270 0000000000000000 ffff8b8af006c270 0000000000000000 : nt!IofCallDriver+0x55
ffffe70fa81488f0 fffff8051bfe454f : ffffd68900000000 ffffe70fa8148b80 0000000c29bff630 ffffe70fa8148b80 : nt!IopSynchronousServiceTail+0x1a8
ffffe70fa8148990 fffff8051be074b8 : ffff8b8aeedac080 0000000000000000 0000000000000000 0000017208c3c798 : nt!NtWriteFile+0x66f
ffffe70fa8148a90 00007fff4638c724 : 0000000000000000 0000000000000000 0000000000000000 0000000000000000 : nt!KiSystemServiceCopyEnd+0x28
0000000c29bff5b8 0000000000000000 : 0000000000000000 0000000000000000 0000000000000000 0000000000000000 : 0x00007fff`4638c724


SYMBOL_NAME: Npfs+d834

MODULE_NAME: Npfs

IMAGE_NAME: Npfs.SYS

IMAGE_VERSION: 10.0.19041.1030

STACK_COMMAND: .thread ; .cxr ; kb

BUCKET_ID_FUNC_OFFSET: d834

FAILURE_BUCKET_ID: AV_INVALID_Npfs!unknown_function

OS_VERSION: 10.0.19041.1

BUILDLAB_STR: vb_release

OSPLATFORM_TYPE: x64

OSNAME: Windows 10

FAILURE_ID_HASH: {5997191b-9534-6e03-7650-35d9877524bb}

Followup: MachineOwner
---------

Please, help me with the issue.
Here's is the link to the first post I opened because of the BSODs. The PC stopped crashing for few days, so I assumed that it's fixed, but then it started crashing again. The memtest didn't find any errors.

 

[blablaalb]

I've got a question regarding the memtest86+: the ram sticks fail within first hour of test if I enable SMP option, but without SMP enabled they were running for 10 hours with 4 passes and no errors.

Why do they pass tests with SMP disabled?
I ran Prime95 with Large FFTs option and it reported the following errors

[Fri Mar  5 13:01:51 2021]
FATAL ERROR: Rounding was 0.4991662188, expected less than 0.4
Hardware failure detected, consult stress.txt file.
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.

If I run Prime95 with the Small FFTs options then no problems are found:

 

[Vic 40]

Might be that xmp messes up the timings somehow, already said to try and set them manually, speed as well.

Maybe the memory controller on the cpu isn't great, maybe the motherboards bios isn't great or the motherboard itself.

No way of testing with another kit?

 

[blablaalb]

Might be that xmp messes up the timings somehow, already said to try and set them manually, speed as well.

Maybe the memory controller on the cpu isn't great, maybe the motherboards bios isn't great or the motherboard itself.

No way of testing with another kit?

I set them manually, both speed and timings.
I wish I had another RAM stick just to test if it's the RAM issue.
What surprises me is that Memtest doesn't recognize memory type and chipset. It says "Unknown" for both of them. Speccy doesn't recognize memory type either. Is that common?

 

[Vic 40]

When? now or earlier on already? And how is that working? If it is not working think about either returning this kit or sell it and get a more compatibel kit. Think it's just some compatibility problem.

 

[blablaalb]

When? now or earlier on already? And how is that working? If it is not working think about either returning this kit or sell it and get a more compatibel kit. Think it's just some compatibility problem.

I set them manually right before testing with Prime95.
It hasn't crashed for last 2 days, but I think It will crash later or sooner. It's been like this before: it didn't crash for a while, so I thought the problem has gone, but then suddenly started crashing again. I remember tweaking with timings and speed settings even before you suggested doing so. I did so many things in order to fix the issue, I've lost the count.

 

[Vic 40]

Well you can just continue with this kit at stock settings or take my advice and test with another.

 

[blablaalb]

Well you can just continue with this kit at stock settings or take my advice and test with another.

I'd love to test with another RAM, but I don't have that option. I'll probably buy another RAM kit later, when i get such opportunity. Thank you for your input.

 

[Vic 40]

You're welcome. Is hard to say more without having another kit to test other than maybe add voltage to the memory controller. Might be Vcore NB ofset voltage, see if you can go +0.05, see if that helps. Could also be the VDDP/SOC voltage, if finding that could try 1.1V, not going over 1.2V. These voltage should you find in the OC Tweaker.

If your German is adequate could you watch this, but maybe that the keywords and vid itself can be helpfull,

View: https://www.youtube.com/watch?v=ygUFGF_FiuA

 

[blablaalb]

Might be good to get another kit to test with.

I bought a new pair of new RAM sticks. Crucial 8GB DDR4-2666 UDIMM CT8G4DFRA266
OCCT and memtest still find errors

I also noticed that no matter what ram stick I have installed the Failing Address is almost the same. If you look at my old post with previous RAM sticks installed it had almost identical Failing Address list. The Err-Bits, Bad and CPU columns also match, no matter what RAM sticks I have installed.

 

[geofelt]

Apparently, you bought two supposedly identical ram sticks.
They are not identical.
Ram is sold in kits for a reason.
Proper operation is only assured for a single kit.

A motherboard must manage all the ram using the same specs of voltage, cas and speed.
The internal workings are designed for the capacity of the kit.
Ram from the same vendor and part number can be made up of differing manufacturing components over time.
Some motherboards, can be very sensitive to this.

 

[blablaalb]

Apparently, you bought two supposedly identical ram sticks.
They are not identical.
Ram is sold in kits for a reason.
Proper operation is only assured for a single kit.

A motherboard must manage all the ram using the same specs of voltage, cas and speed.
The internal workings are designed for the capacity of the kit.
Ram from the same vendor and part number can be made up of differing manufacturing components over time.
Some motherboards, can be very sensitive to this.

IDK, I did some research before buying sticks and It was said as long as two ram sticks have the same speed, voltage, latency etc. they should work fine. With the previous kit I tested each stick at a time and still had the same results.
Here cpu-z report for both of the sticks

 

[geofelt]

The operative word is "supposedly"
Most of the time, disparate sticks with the same specs work just fine.
Particularly with intel motherboards.
Ryzen, on the other hand is very tightly tied to ram for performance and functionality.
You may have other underlying issues that are causing you problem.

Since the ram failed memtest, I would try to return it in favor of a kit that is explicitly supported for your motherboard on the ram QVL list.
There are different versions, depending on the processor used; pick the right list.

 

[blablaalb]

Since the ram failed memtest, I would try to return it in favor of a kit that is explicitly supported for your motherboard on the ram QVL list.

I have only one question. What if they retest the ram sticks on their side and they won't fail, they might refuse me to replace those sticks with different ones? If I bought the wrong sticks then it's my problem, they won't exchange them for different ones.

 

[Herr B]

I have had many BSOD experiences with tools such Driver easy. BSODS most of the times hint to driver issues.

1. Load stock settings in Bios (to evade all possibly incorrect settings here)
2. Fresh windows install formatting the harddrive (to evade all possibly corrupt driver installations - removing old drivers does not always work)
3. Get drivers from Oem website and install them

That's the first steps I'd do as an IT technician in order to make sure all drivers are set up properly and I do not have to deal with any faulty residue from previous experiments.

Most oems webpages are having the very same navigation leading to the drivers:
Manufacturers Webpage -> support -> Drivers and downloads

Once you did that you can be fairly sure there is a hardware issue.

If possible, send link to minidump file. Minidump contains the basic bsod information. Its normally located in C:\Windows\minidump

 

[geofelt]

I have only one question. What if they retest the ram sticks on their side and they won't fail, they might refuse me to replace those sticks with different ones? If I bought the wrong sticks then it's my problem, they won't exchange them for different ones.

If you bought directly from crucial, they might well test the sticks and find them good individually.
They would be returned to you and you could sell them yourself.
If you bought from a retailer, the defective return policy will be in effect and the retailer should refund your money. They will not have the means to test every return.
A retailer is likely to look favorably if you offer to buy a replacement kit.

All of this assumes that your problems are not rooted elsewhere.

 

[Herr B]

Agree. Vendors usually do that whithout any questions.
By the fact that you bought different sticks already kind of rules that error out to me. The fact that the two sticks should not match well can be easily ruled out by just testing one stick at a time.

 

[blablaalb]

By the fact that you bought different sticks already kind of rules that error out to me.

They are actually the same sticks; the same part number, frequency, timings. etc. They are identical sticks, just weren't being sold as a single kit. Before buying those sticks I had the same problems with another kit of ram sticks.

 

[blablaalb]

I just rechecked the QVL list and the previous kit of RAMs I had installed is listed there. (CMK16GX4M2A2400C16) IF it's not because of the ram then what else it might be?

 

[Herr B]

I have had many BSOD experiences with tools such Driver easy. BSODS most of the times hint to driver issues.

1. Load stock settings in Bios (to evade all possibly incorrect settings here)
2. Fresh windows install formatting the harddrive (to evade all possibly corrupt driver installations - removing old drivers does not always work)
3. Get drivers from Oem website and install them

That's the first steps I'd do as an IT technician in order to make sure all drivers are set up properly and I do not have to deal with any faulty residue from previous experiments.

Most oems webpages are having the very same navigation leading to the drivers:
Manufacturers Webpage -> support -> Drivers and downloads

Once you did that you can be fairly sure there is a hardware issue.

If possible, send link to minidump file. Minidump contains the basic bsod information. Its normally located in C:\Windows\minidump

 

[alexbirdie]

Any chance to try another PSU?

 

[blablaalb]

Any chance to try another PSU?

Unfortunately don't have another. I bought the PSU on ebay from newegg. The PSU is Thermaltake Toughpower GX1 RGB 500W ATX 12V v2.4 and EPS v2.92 80 PLUS GOLD Cert it had pretty good reviews for its price and it's gold-certified.

 

[Vic 40]

Your motherboard requires a bios update to level P2.0 to support your processor.
https://www.asrock.com/mb/AMD/B450M Pro4-F/#CPU

'Something odd about this, it seems to be necessary to update to at least 2.00 to make the 1600 12nm work, but at the same time they say not to update to it when using pinnacle ridge which the 1600 12nm is .....

Maybe OP should ask support of Asrock what the hell that means. Don't know if he has upgrade the bios above this 2.00 which likely should be better for ram compatibility.

Could also be that the motehrboard just doesn't work right or even the memory controller on the cpu when having faults in memtest86. Although even memtest itself can get it wrong. You tested the sticks without overclock/ XMP enabled?

I set voltage to 1.3v and it made the system even more unstable.

Sure that was Dram voltage? Should be fine upt to 1.35V and even 1.4V should work.

 

[blablaalb]

'Something odd about this, it seems to be necessary to update to at least 2.00 to make the 1600 12nm work, but at the same time they say not to update to it when using pinnacle ridge which the 1600 12nm is .....

Maybe OP should ask support of Asrock what the hell that means. Don't know if he has upgrade the bios above this 2.00 which likely should be better for ram compatibility.

Could also be that the motehrboard just doesn't work right or even the memory controller on the cpu when having faults in memtest86. Although even memtest itself can get it wrong. You tested the sticks without overclock/ XMP enabled?



Sure that was Dram voltage? Should be fine upt to 1.35V and even 1.4V should work.

I also was confused by it. Since I already had problems with system stability I updated the BIOS to the latest version, but then downgraded it to the p2.0. The new ram sticks I bought don't have XMP profile.
I'm pretty sure it was DRAM voltage. The VCore option is above ram options and was set to automatic and hadn't touched it. A few days ago I set the VCore Offset to -50 since it was 1.35-1.40v even when system was idle.

It seems like something is wrong with the Ryzen's SMT. With SMT disabled system seems to be more stable. Also the memtest: it doesn' find any errors even with the old ram sticks unless I enable the multithreading. The old ram sticks were tested with memtest for 9 and half hours without multithreading enabled and no errors were found, but if enable the multithreading option (F2) in the memtest then it finds errors within the first hour of running, sometimes within few minutes. I disabled the SMT option in the BIOS and let the memtest run for 2 complete passes with multithreading enabled (F2). It utilized all 6 physical cores and didn't find any errors.

 

[Vic 40]

Might just have a bad cpu, would try and rma.

 

[blablaalb]

I disabled the PSS Support in the UEFI and it's been a few months already since I had BSOD last time. The pc doesn't crash anymore and it passes all the memory tests.
The PSS is power-saving feature and it was previously called Cool and Quiet. I googled a bit, it was reported on several websites to cause system instabilities.
I think this thread can be closed. Thanks to everyone who tried to help me. Cheers! 🥳