XMP Profile 1 not stable on Kingston Fury Renegade 7200

[mrpkv]

Hello,

 

I have the following configuration:

 

ASUS ROG MAXIMUS HERO Z790

Intel i9-13900K

32GB(2x16) Kingston Fury Renegade at 7200MHz

 

With XMP I and Profile 1 enabled at 7200MHz, MCE disabled, CPU limited at 253W I get application crashes a lot (eg. Steam, Microsoft Office Apps, Chrome STATUS_ACCES_VIOLATION, a lot of Windows Explorer crashes and crashes in games after some time of playing). Also, during the AIDA64 system stability test I get hardware failure after 30 seconds and BSODs with error code MEMORY_MANAGEMENT. I must mention that running Windows Memory Diagnostic or MemTest86 reports 0 errors, I have done these tests multiple times. 
 

When I lower the frequency from 7200MHz to 6933MHz but keep the same timings and voltages I notice improvements, AIDA64 system stability test does not report hardware failure anymore and the crashes and BSODs are less frequent, but they still exist. OCCT memory test also runs fine for 15 minutes at 6933MHz, did not run it for longer. 
 

It also feels like a matter of luck sometimes, as I had days when I saw 0 errors in the reliability monitor, but the next day Windows Explorer or Steam start crashing as soon as I start up the PC and I need to restart otherwise the keep crashing in a loop. 

For a couple of weeks I’ve been trying to figure what is the reason for my system being unstable.

 

I don’t think there is a problem with the CPU or memory, I think there is a setting problem with the way I configure the BIOS or the drivers are problematic as I let ASUS ARMOURY CRATE, NVIDA GeForce Experience and Windows Update install my drivers. 
 

Thank you!

[Radium_Angel]

List XMP speeds of RAM are under "ideal" (laboratory) conditions and if you are lucky, you get them. If not...then you don't.

Sounds like the RAM you bought can't reach the speeds are you are trying to achieve 

[191x7]

The official ddr5 support for the i9 13900K is 5600 and anything above that is pure luck.

[RONOTHAN##]

As someone who has spent a lot of time overclocking DDR5 as of recently, speeds above ~7000MT/s are super difficult to get stable. 7200MT/s kits do have pretty good performance, but at the same time they are a ton of work to get stable, and just enabling XMP isn't really enough to get them to work a lot of the time. 

 

Usually the issue that arises is memory controller instability. A good stress test that seems to pick up on those errors very well is Y-Cruncher's VST test, if the system is blue screening that frequently it should crash within the first test with the setup you've got. If you do want to try and get it stable at 7200MT/s, your best bet is to focus on that VST test. There are 3 main voltage settings that have a bit affect on the memory controller's stability: VDDQ TX, VDD2, and VCCSA. ASUS does call them different things in the BIOS (ASUS just has the tendency to do this for some reason), so you want to be looking for IIRC IVR Transmitter voltage, Memory Controller voltage, and System Agent voltage. All three of those voltage tend to sweet spot, and that sweet spot changes depending on the BIOS revision you're on, memory speed your at, and to some extent each other. You basically want to manipulate each one of those voltages until you start getting it stable, trying to keep IVR Transmitter and System Agent below 1.45V (though the sweet spot should happen below that mark), and I'd be trying to stay below 1.8V on the memory controller voltage (though again, you really shouldn't be getting anywhere near this for where the sweet spot should be). I don't really have any experience on ASUS boards, so I can't tell you what they like and what they don't for those voltage rails, though going off of screen shots I've seen something like 1.4V on IVR Transmitter and Memory controller with 1.3V for VCCSA should be about where each of those voltage sweet spots if you want a starting point (though again, this is dependent on a lot of factors, and even to some degree the exact board you got from the factory can behave differently from one to the next). 

 

The other option is to just return the kit and go for a 6400MT/s rated kit, which realistically is what I'd be recommending. At this point they're both Hynix A die based, so in the event that you do want to run a 7200MT/s kit it should overclock to that point, and since a 7200MT/s rated kit is fast enough that you need to do manual voltage tuning anyway (the actual hard part of DDR5 tuning) I don't really see any downsides to it since you'll be saving money, getting a kit that's got an XMP that will actually work, and if you decide to put in the effort to tune the memory controller in the future you still can run 7200MT/s without much difficulty. 

[mrpkv]

4 hours ago, RONOTHAN## said:

As someone who has spent a lot of time overclocking DDR5 as of recently, speeds above ~7000MT/s are super difficult to get stable. 7200MT/s kits do have pretty good performance, but at the same time they are a ton of work to get stable, and just enabling XMP isn't really enough to get them to work a lot of the time. 

 

Usually the issue that arises is memory controller instability. A good stress test that seems to pick up on those errors very well is Y-Cruncher's VST test, if the system is blue screening that frequently it should crash within the first test with the setup you've got. If you do want to try and get it stable at 7200MT/s, your best bet is to focus on that VST test. There are 3 main voltage settings that have a bit affect on the memory controller's stability: VDDQ TX, VDD2, and VCCSA. ASUS does call them different things in the BIOS (ASUS just has the tendency to do this for some reason), so you want to be looking for IIRC IVR Transmitter voltage, Memory Controller voltage, and System Agent voltage. All three of those voltage tend to sweet spot, and that sweet spot changes depending on the BIOS revision you're on, memory speed your at, and to some extent each other. You basically want to manipulate each one of those voltages until you start getting it stable, trying to keep IVR Transmitter and System Agent below 1.45V (though the sweet spot should happen below that mark), and I'd be trying to stay below 1.8V on the memory controller voltage (though again, you really shouldn't be getting anywhere near this for where the sweet spot should be). I don't really have any experience on ASUS boards, so I can't tell you what they like and what they don't for those voltage rails, though going off of screen shots I've seen something like 1.4V on IVR Transmitter and Memory controller with 1.3V for VCCSA should be about where each of those voltage sweet spots if you want a starting point (though again, this is dependent on a lot of factors, and even to some degree the exact board you got from the factory can behave differently from one to the next). 

 

The other option is to just return the kit and go for a 6400MT/s rated kit, which realistically is what I'd be recommending. At this point they're both Hynix A die based, so in the event that you do want to run a 7200MT/s kit it should overclock to that point, and since a 7200MT/s rated kit is fast enough that you need to do manual voltage tuning anyway (the actual hard part of DDR5 tuning) I don't really see any downsides to it since you'll be saving money, getting a kit that's got an XMP that will actually work, and if you decide to put in the effort to tune the memory controller in the future you still can run 7200MT/s without much difficulty. 

Thank you very much for the detailed and very informative answer.

 

I will most probably swap it for a 6400 kit, but for now I need to make this work as I need to finish a project.

 

I tried to use the second profile of XMP at 6800MHz which has different timings and voltages and it BSODed 10 seconds into Windows.

 

I reverted to the first profile of XMP at 7200MHz and kept the timings and voltages but lowered the frequency to 6800MHz. So far it feels stable, although I did not use the Y-Cruncher VST test yet.

 

My CPU is also undervolted by 95mV, could this influence it also?

[RONOTHAN##]

18 minutes ago, mrpkv said:

My CPU is also undervolted by 95mV, could this influence it also?

Technically possible, but that's only really if your undervolt wasn't really stable to begin with. I'm gonna stick with saying that your CPU's memory controller voltages aren't set right by the BIOS with auto settings and that's what's causing the instability. It could also be that you've got a CPU with a very weak IMC where 7200MT/s is about where it tops out, and running chips at their limits takes a lot of fine tuning, or that your chip has very weird voltage requirements.

[Mark Kaine]

4 hours ago, mrpkv said:

for now I need to make this work as I need to finish a project.

 

have you tried to just not run any XMP/overclocking? 

 

i mean if that still doesn't work then there are probably more issues... but it should usually work

[mrpkv]

14 minutes ago, Mark Kaine said:

have you tried to just not run any XMP/overclocking? 

 

i mean if that still doesn't work then there are probably more issues... but it should usually work

Did not try with XMP turned off yet, this will be my next step. 
 

4 hours ago, mrpkv said:

I reverted to the first profile of XMP at 7200MHz and kept the timings and voltages but lowered the frequency to 6800MHz.

Following what I said earlier, this is not stable either, no BSOD but applications keep crashing. 
 

4 hours ago, RONOTHAN## said:

Technically possible, but that's only really if your undervolt wasn't really stable to begin with.

The undervolt feels stable, did not notice anything unusual. 

[Mark Kaine]

31 minutes ago, mrpkv said:

The undervolt feels stable, did not notice anything unusual. 

except the crashing heh?

 

you really should reset CMOS and leave everything at default right now, its entirely possible your undervolt isn't stable, and your cpu memory controller has issues with the RAM XMP on top of it (because that's what this all sounds like)

[mrpkv]

1 hour ago, Mark Kaine said:

except the crashing heh?

 

you really should reset CMOS and leave everything at default right now, its entirely possible your undervolt isn't stable, and your cpu memory controller has issues with the RAM XMP on top of it (because that's what this all sounds like)

The crashes were happening before the undervolt. After I undervolted the CPU I performed 30-minute runs in Cinebench R23 (getting between 39500-40500) and AIDA64 stability tests (15-30 minutes also), no BSOD, no hardware failure. All of these with XMP Profile 1 but lowered frequency from 7200MHz to 6933MHz. Felt stable for some time then application crashes begun and eventually BSODs, all randomly, did not matter if it was under intensive load or not. 
 

The most recent BSOD error codes were: CRITICAL_PROCESS_DIED (svchost.exe), IRQL_NOT_LESS_OR_EQUAL and SYSTEM_SERVICE_EXCEPTION.

 

Could this happen because of drivers also? I was thinking of reinstalling Windows and properly installing the drivers without any external application. 
 

It confuses me that memory tests show no errors with XMP enabled and benchmarks or stability tests do not crash the system.

[RONOTHAN##]

3 hours ago, mrpkv said:

Could this happen because of drivers also? I was thinking of reinstalling Windows and properly installing the drivers without any external application. 

It's possible, but those error codes are more common among memory errors than driver errors. Try running JEDEC to see if that fixes it. 

 

3 hours ago, mrpkv said:

It confuses me that memory tests show no errors with XMP enabled and benchmarks or stability tests do not crash the system.

The memory tests you've run are also not the most sensitive to memory errors. There are unstable OCs I've run that pass both MemTest86 and Aida64. Besides, this is likely an IMC issue, which both of those tests don't really show instability from (TestMem5 does to some degree, though VST does a lot). 

[mrpkv]

10 hours ago, RONOTHAN## said:

It's possible, but those error codes are more common among memory errors than driver errors. Try running JEDEC to see if that fixes it. 

 

The memory tests you've run are also not the most sensitive to memory errors. There are unstable OCs I've run that pass both MemTest86 and Aida64. Besides, this is likely an IMC issue, which both of those tests don't really show instability from (TestMem5 does to some degree, though VST does a lot). 

Switched to JEDEC, 10 hours of no application crashes or BSOD.

 

Also, I tried to run the Y cruncher VST test but I do not know how to activate it, it is not listed under test names, first time using this. Could you tell me?

[RONOTHAN##]

11 minutes ago, mrpkv said:

Switched to JEDEC, 10 hours of no application crashes or BSOD.

 

Also, I tried to run the Y cruncher VST test but I do not know how to activate it, it is not listed under test names, first time using this. Could you tell me?

There are two ways of running it

1. Opening up command prompt and navigating to the folder containing the Y cruncher exe file. Enter the following command

   y-cruncher stress vst

2. Double click the executable, press 1 (stress), press 8 (clear all stress tests), press 18 (VST), and press 0 (start test). I'm going off memory for this, so I might be slightly wrong about the specific numbers, but I'm pretty sure they're correct. 

[mrpkv]

47 minutes ago, RONOTHAN## said:

There are two ways of running it

1. Opening up command prompt and navigating to the folder containing the Y cruncher exe file. Enter the following command

   y-cruncher stress vst

2. Double click the executable, press 1 (stress), press 8 (clear all stress tests), press 18 (VST), and press 0 (start test). I'm going off memory for this, so I might be slightly wrong about the specific numbers, but I'm pretty sure they're correct. 

Thank you! It works and it succeeded the VST test at JEDEC. Does that tell me that there is no problem with the CPU and RAM? I mean some factory defects.

 

Also, how do I get Y cruncher to show me the options for 8 and 18 (as you mentioned me to press), I only see those and the list of tests from below does not contain anything:

 

[RONOTHAN##]

4 minutes ago, mrpkv said:

Thank you! It works and it succeeded the VST test at JEDEC. Does that tell me that there is no problem with the CPU and RAM? I mean some factory defects.

 

Also, how do I get Y cruncher to show me the options for 8 and 18 (as you mentioned me to press), I only see those and the list of tests from below does not contain anything:

 

OK, that's weird. This is what it should look like:

 

I'd redownload it to see if there's something that broke when downloading it or extracting it. Also, where did you get the download link, since you should've gotten it from here

[mrpkv]

19 minutes ago, RONOTHAN## said:

OK, that's weird. This is what it should look like:

 

I'd redownload it to see if there's something that broke when downloading it or extracting it. Also, where did you get the download link, since you should've gotten it from here

I used the same download link, but I figured I was not running Y cruncher as administrator. After running as administrator it looks like in your screenshot.

 

Retried the stability test with the 6 tests BKT BBP SFT N64 HNT VST enabled and I got a blue screen during SFT.

 

IRQL_NOT_LESS_OR_EQUAL (a)
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 a kernel debugger is available get the stack backtrace.
Arguments:
Arg1: 0000000000000000, memory referenced
Arg2: 0000000000000002, IRQL
Arg3: 0000000000000000, bitfield :
	bit 0 : value 0 = read operation, 1 = write operation
	bit 3 : value 0 = not an execute operation, 1 = execute operation (only on chips which support this level of status)
Arg4: fffff8052685fe4d, address which referenced memory

 

CUSTOMER_CRASH_COUNT:  1

PROCESS_NAME:  chrome.exe

TRAP_FRAME:  ffffb40124fd71e0 -- (.trap 0xffffb40124fd71e0)
NOTE: The trap frame does not contain all registers.
Some register values may be zeroed or incorrect.
rax=0000000000000013 rbx=0000000000000000 rcx=7fb300be380f0000
rdx=000000000012aaad rsi=0000000000000000 rdi=0000000000000000
rip=fffff8052685fe4d rsp=ffffb40124fd7370 rbp=ffffb40124fd73d0
 r8=ffffb65b11d20068  r9=0000474801a12000 r10=0000000000000000
r11=ffffdb0000000000 r12=0000000000000000 r13=0000000000000000
r14=0000000000000000 r15=0000000000000000
iopl=0         nv up ei pl nz na pe nc
nt!MiAllocateWsle+0x58d:
fffff805`2685fe4d 0008            add     byte ptr [rax],cl ds:00000000`00000013=??
Resetting default scope

MISALIGNED_IP: 
nt!MiAllocateWsle+58d
fffff805`2685fe4d 0008            add     byte ptr [rax],cl

STACK_TEXT:  
ffffb401`24fd7098 fffff805`26a3e2a9     : 00000000`0000000a 00000000`00000000 00000000`00000002 00000000`00000000 : nt!KeBugCheckEx
ffffb401`24fd70a0 fffff805`26a39934     : 00000000`00000001 00000000`00000000 00000000`00000202 00000000`00000000 : nt!KiBugCheckDispatch+0x69
ffffb401`24fd71e0 fffff805`2685fe4d     : ffff8c8a`d8fae740 00004748`00000000 00000000`00000002 00004748`01a12000 : nt!KiPageFault+0x474
ffffb401`24fd7370 fffff805`2685f1df     : ffff8c8a`d8fae740 ffffb623`a400d090 ffffdb00`0b621340 8a000003`cb5bc867 : nt!MiAllocateWsle+0x58d
ffffb401`24fd7410 fffff805`2685e9e4     : ffffb401`24fd7720 00000000`00000000 00000000`00000001 00000000`00000011 : nt!MiCompletePrivateZeroFault+0x77f
ffffb401`24fd7590 fffff805`2685c8f8     : ffffb401`24fd7720 00000000`00000000 00000000`00000000 ffff8c8a`d8fae740 : nt!MiResolvePrivateZeroFault+0x7e4
ffffb401`24fd76d0 fffff805`26859310     : ffffb401`24fd7820 ffffb623`a400d090 00000000`00000080 00000000`00000002 : nt!MiResolveDemandZeroFault+0x448
ffffb401`24fd77c0 fffff805`26857442     : 00000000`00000111 00000000`00000004 00000000`c0000016 00000000`0021a000 : nt!MiDispatchFault+0x360
ffffb401`24fd7900 fffff805`26a39829     : ffff8c8a`cccd9080 ffff8c8a`d9a4f5c0 00004748`04abb580 ffff8c8a`00000000 : nt!MmAccessFault+0x152
ffffb401`24fd7a20 00007ffe`474a7585     : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : nt!KiPageFault+0x369
00000091`231fb8a0 00000000`00000000     : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : 0x00007ffe`474a7585


SYMBOL_NAME:  nt!MiAllocateWsle+58d

IMAGE_VERSION:  10.0.22621.1194

STACK_COMMAND:  .cxr; .ecxr ; kb

IMAGE_NAME:  GenuineIntel.sys

MODULE_NAME: GenuineIntel

FAILURE_BUCKET_ID:  IP_MISALIGNED_GenuineIntel.sys

OSPLATFORM_TYPE:  x64

OSNAME:  Windows 10

FAILURE_ID_HASH:  {e930d917-b247-3f7f-23cb-88a0c9f1b274}

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

 

It says the process name is chrome.exe, it was open during the test run.

[mrpkv]

10 minutes ago, mrpkv said:

I used the same download link, but I figured I was not running Y cruncher as administrator. After running as administrator it looks like in your screenshot.

 

Retried the stability test with the 6 tests BKT BBP SFT N64 HNT VST enabled and I got a blue screen during SFT.

 

IRQL_NOT_LESS_OR_EQUAL (a)
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 a kernel debugger is available get the stack backtrace.
Arguments:
Arg1: 0000000000000000, memory referenced
Arg2: 0000000000000002, IRQL
Arg3: 0000000000000000, bitfield :
	bit 0 : value 0 = read operation, 1 = write operation
	bit 3 : value 0 = not an execute operation, 1 = execute operation (only on chips which support this level of status)
Arg4: fffff8052685fe4d, address which referenced memory

 

CUSTOMER_CRASH_COUNT:  1

PROCESS_NAME:  chrome.exe

TRAP_FRAME:  ffffb40124fd71e0 -- (.trap 0xffffb40124fd71e0)
NOTE: The trap frame does not contain all registers.
Some register values may be zeroed or incorrect.
rax=0000000000000013 rbx=0000000000000000 rcx=7fb300be380f0000
rdx=000000000012aaad rsi=0000000000000000 rdi=0000000000000000
rip=fffff8052685fe4d rsp=ffffb40124fd7370 rbp=ffffb40124fd73d0
 r8=ffffb65b11d20068  r9=0000474801a12000 r10=0000000000000000
r11=ffffdb0000000000 r12=0000000000000000 r13=0000000000000000
r14=0000000000000000 r15=0000000000000000
iopl=0         nv up ei pl nz na pe nc
nt!MiAllocateWsle+0x58d:
fffff805`2685fe4d 0008            add     byte ptr [rax],cl ds:00000000`00000013=??
Resetting default scope

MISALIGNED_IP: 
nt!MiAllocateWsle+58d
fffff805`2685fe4d 0008            add     byte ptr [rax],cl

STACK_TEXT:  
ffffb401`24fd7098 fffff805`26a3e2a9     : 00000000`0000000a 00000000`00000000 00000000`00000002 00000000`00000000 : nt!KeBugCheckEx
ffffb401`24fd70a0 fffff805`26a39934     : 00000000`00000001 00000000`00000000 00000000`00000202 00000000`00000000 : nt!KiBugCheckDispatch+0x69
ffffb401`24fd71e0 fffff805`2685fe4d     : ffff8c8a`d8fae740 00004748`00000000 00000000`00000002 00004748`01a12000 : nt!KiPageFault+0x474
ffffb401`24fd7370 fffff805`2685f1df     : ffff8c8a`d8fae740 ffffb623`a400d090 ffffdb00`0b621340 8a000003`cb5bc867 : nt!MiAllocateWsle+0x58d
ffffb401`24fd7410 fffff805`2685e9e4     : ffffb401`24fd7720 00000000`00000000 00000000`00000001 00000000`00000011 : nt!MiCompletePrivateZeroFault+0x77f
ffffb401`24fd7590 fffff805`2685c8f8     : ffffb401`24fd7720 00000000`00000000 00000000`00000000 ffff8c8a`d8fae740 : nt!MiResolvePrivateZeroFault+0x7e4
ffffb401`24fd76d0 fffff805`26859310     : ffffb401`24fd7820 ffffb623`a400d090 00000000`00000080 00000000`00000002 : nt!MiResolveDemandZeroFault+0x448
ffffb401`24fd77c0 fffff805`26857442     : 00000000`00000111 00000000`00000004 00000000`c0000016 00000000`0021a000 : nt!MiDispatchFault+0x360
ffffb401`24fd7900 fffff805`26a39829     : ffff8c8a`cccd9080 ffff8c8a`d9a4f5c0 00004748`04abb580 ffff8c8a`00000000 : nt!MmAccessFault+0x152
ffffb401`24fd7a20 00007ffe`474a7585     : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : nt!KiPageFault+0x369
00000091`231fb8a0 00000000`00000000     : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : 0x00007ffe`474a7585


SYMBOL_NAME:  nt!MiAllocateWsle+58d

IMAGE_VERSION:  10.0.22621.1194

STACK_COMMAND:  .cxr; .ecxr ; kb

IMAGE_NAME:  GenuineIntel.sys

MODULE_NAME: GenuineIntel

FAILURE_BUCKET_ID:  IP_MISALIGNED_GenuineIntel.sys

OSPLATFORM_TYPE:  x64

OSNAME:  Windows 10

FAILURE_ID_HASH:  {e930d917-b247-3f7f-23cb-88a0c9f1b274}

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

 

It says the process name is chrome.exe, it was open during the test run.

Repeated the test twice with no background applications, same BSOD 15 seconds into the SFT test.

[RONOTHAN##]

6 minutes ago, mrpkv said:

Repeated the test twice with no background applications, same BSOD 15 seconds into the SFT test.

Are you still undervolt Ed? SFT is very CPU heavy and will crash any sketchy overclock

[mrpkv]

23 minutes ago, RONOTHAN## said:

Are you still undervolt Ed? SFT is very CPU heavy and will crash any sketchy overclock

Yes, I was still undervolted. After disabling the undervolt SFT passed.

 

Tried the XMP Profile 3 at 6400MHz and ran all the tests and they passed:

 

 

Can I judge this as stable? Or should I keep running the tests for longer?

[RONOTHAN##]

12 minutes ago, mrpkv said:

Yes, I was still undervolted. After disabling the undervolt SFT passed.

 

Tried the XMP Profile 3 at 6400MHz and ran all the tests and they passed:

 

 

Can I judge this as stable? Or should I keep running the tests for longer?

I'd run the tests for longer, and since you have disabled the undervolt I'd run just VST for a while on the 6400MT/s profile to make sure that's stable. Personally I try to have my systems pass VST for 1 hour to declare them stable

[mrpkv]

On 2/8/2023 at 11:04 PM, RONOTHAN## said:

I'd run the tests for longer, and since you have disabled the undervolt I'd run just VST for a while on the 6400MT/s profile to make sure that's stable. Personally I try to have my systems pass VST for 1 hour to declare them stable

Thank you very much for the help!

 

It seems stable now, no BSOD after 2 days of operating the PC, VST ran for 2 hours and passed, OCCT also. 

 

I still get some random application crashes which I cannot explain, for example Adobe Premiere Pro crashed while editing and Hogwarts Legacy during loading, weirdly enough I did not find anything in the Event Viewer or Reliability Monitor to see what happened. I was thinking maybe the Windows got broken because of so many BSODs during the last weeks (I tried sfc /scannow and DISM RestoreHealth and AnalyzeComponents), what do you think? If this gets annoying I am pondering to reinstall Windows with fresh drivers.

[RONOTHAN##]

3 minutes ago, mrpkv said:

I still get some random application crashes which I cannot explain, for example Adobe Premiere Pro crashed while editing and Hogwarts Legacy during loading

To be fair, both of those programs are pretty well known for being unstable in general, so it could just be that they crashed. 

 

4 minutes ago, mrpkv said:

I was thinking maybe the Windows got broken because of so many BSODs during the last weeks (I tried sfc /scannow and DISM RestoreHealth and AnalyzeComponents), what do you think?

It can happen, I've bricked Windows installs with unstable memory overclocks before (the worst one was when I was trying to overclock OEM Samsung B die DDR4, I broke the OS at 3200 CL16 because those DIMMs were so terrible). For the most part if the system can get into Windows and there are no obvious issues immediately, it's probably OK, but if you start noticing odd behavior it's not a bad idea to try. 

[mrpkv]

On 2/11/2023 at 3:20 AM, RONOTHAN## said:

To be fair, both of those programs are pretty well known for being unstable in general, so it could just be that they crashed. 

 

It can happen, I've bricked Windows installs with unstable memory overclocks before (the worst one was when I was trying to overclock OEM Samsung B die DDR4, I broke the OS at 3200 CL16 because those DIMMs were so terrible). For the most part if the system can get into Windows and there are no obvious issues immediately, it's probably OK, but if you start noticing odd behavior it's not a bad idea to try. 

All went fine until today when randomly browsing the web I got a PFN_LIST_CORRUPT BSOD. I have not reinstalled the Windows yet.

 

I do not know what to do anymore, maybe I should try another kit of RAM? I was thinking to upgrade to 64GBs also, I checked the QVL and found that all the kits with higher XMP frequency than 6000MHz are DS. Do you think SS kits are more stable than DS kits?

 

I am thinking of ordering the G.SKILL D564GB 6000-30 Trident Z5 RGB BK K2 GSK | F5-6000J3040G32GX2-TZ5RK, what is your opinion?

 

*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

PFN_LIST_CORRUPT (4e)
Typically caused by drivers passing bad memory descriptor lists (ie: calling
MmUnlockPages twice with the same list, etc).  If a kernel debugger is
available get the stack trace.
Arguments:
Arg1: 0000000000000002, A list entry was corrupt
Arg2: 00000000003666eb, entry in list being removed
Arg3: 000000000089f7ff, highest physical page number
Arg4: 0000000000000001, reference count of entry being removed

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


KEY_VALUES_STRING: 1

    Key  : Analysis.CPU.mSec
    Value: 937

    Key  : Analysis.DebugAnalysisManager
    Value: Create

    Key  : Analysis.Elapsed.mSec
    Value: 2565

    Key  : Analysis.IO.Other.Mb
    Value: 0

    Key  : Analysis.IO.Read.Mb
    Value: 0

    Key  : Analysis.IO.Write.Mb
    Value: 0

    Key  : Analysis.Init.CPU.mSec
    Value: 312

    Key  : Analysis.Init.Elapsed.mSec
    Value: 16763

    Key  : Analysis.Memory.CommitPeak.Mb
    Value: 150

    Key  : Bugcheck.Code.DumpHeader
    Value: 0x4e

    Key  : Bugcheck.Code.Register
    Value: 0x4e

    Key  : Dump.Attributes.AsUlong
    Value: 1008

    Key  : Dump.Attributes.DiagDataWrittenToHeader
    Value: 1

    Key  : Dump.Attributes.ErrorCode
    Value: 0

    Key  : Dump.Attributes.KernelGeneratedTriageDump
    Value: 1

    Key  : Dump.Attributes.LastLine
    Value: Dump completed successfully.

    Key  : Dump.Attributes.ProgressPercentage
    Value: 0


FILE_IN_CAB:  021223-11656-01.dmp

DUMP_FILE_ATTRIBUTES: 0x1008
  Kernel Generated Triage Dump

BUGCHECK_CODE:  4e

BUGCHECK_P1: 2

BUGCHECK_P2: 3666eb

BUGCHECK_P3: 89f7ff

BUGCHECK_P4: 1

BLACKBOXBSD: 1 (!blackboxbsd)


BLACKBOXNTFS: 1 (!blackboxntfs)


BLACKBOXPNP: 1 (!blackboxpnp)


BLACKBOXWINLOGON: 1

CUSTOMER_CRASH_COUNT:  1

PROCESS_NAME:  System

STACK_TEXT:  
ffffbc07`e974f598 fffff807`7169874d     : 00000000`0000004e 00000000`00000002 00000000`003666eb 00000000`0089f7ff : nt!KeBugCheckEx
ffffbc07`e974f5a0 fffff807`71487f47     : ffffaf80`0a334c10 00000000`00000001 00000000`00000001 00000000`00000000 : nt!MiUnlinkPageFromListEx+0x1c630d
ffffbc07`e974f780 fffff807`71826122     : fffff807`71e6ae80 00000000`00000076 00000000`00000000 ffffffff`ffffffff : nt!MiReferencePageForModifiedWrite+0x4f
ffffbc07`e974f7b0 fffff807`71826ec3     : 00000000`00000003 ffffbc07`00000000 ffffbc07`e974f920 00000000`00000100 : nt!MiFillNoReservationCluster+0x25a
ffffbc07`e974f810 fffff807`716ed7ec     : ffffe401`b611e010 fffff807`00000004 00000000`00000100 fffff807`71e6b218 : nt!MiGatherPagefilePages+0x31f
ffffbc07`e974f910 fffff807`7140f4a7     : ffffe401`b614f080 ffffe401`b614f080 00000000`00000080 fffff807`715e6b30 : nt!MiModifiedPageWriter+0x106cbc
ffffbc07`e974fb30 fffff807`7162dbb4     : ffff9f01`2ae51180 ffffe401`b614f080 fffff807`7140f450 500bb212`000e3412 : nt!PspSystemThreadStartup+0x57
ffffbc07`e974fb80 00000000`00000000     : ffffbc07`e9750000 ffffbc07`e9749000 00000000`00000000 00000000`00000000 : nt!KiStartSystemThread+0x34


SYMBOL_NAME:  nt!MiUnlinkPageFromListEx+1c630d

MODULE_NAME: nt

IMAGE_VERSION:  10.0.22621.1194

STACK_COMMAND:  .cxr; .ecxr ; kb

IMAGE_NAME:  ntkrnlmp.exe

BUCKET_ID_FUNC_OFFSET:  1c630d

FAILURE_BUCKET_ID:  0x4E_2_nt!MiUnlinkPageFromListEx

OSPLATFORM_TYPE:  x64

OSNAME:  Windows 10

FAILURE_ID_HASH:  {ac440f38-61f7-1e91-d444-bdc71249d092}

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

 

[RONOTHAN##]

1 hour ago, mrpkv said:

I was thinking to upgrade to 64GBs also, I checked the QVL and found that all the kits with higher XMP frequency than 6000MHz are DS. Do you think SS kits are more stable than DS kits?

So generally that refers to the organization of the memory sticks. DS means dual sided or dual rank memory, while SS means single sided or single rank memory. Currently, any 32GB DIMM is going to be dual rank just because it's the only way possible with the available memory tech currently. As a rule, more memory ranks in a system means the harder the memory controller is working, though more ranks does have a performance per clock advantage (that was a bigger deal on DDR4 though than it is on DDR5, rank interleaving on DDR5 doesn't really add that much, but it does add something). So yes, SS sticks are going to be a little more stable with no tuning since they're easier to run, though at 6000MT/s there really shouldn't be all that much difference. 

 

1 hour ago, mrpkv said:

I am thinking of ordering the G.SKILL D564GB 6000-30 Trident Z5 RGB BK K2 GSK | F5-6000J3040G32GX2-TZ5RK, what is your opinion?

 

It's a fine kit, though if you don't really need 64GB of RAM I would personally stick to the current kit you've got. 

 

1 hour ago, mrpkv said:

All went fine until today when randomly browsing the web I got a PFN_LIST_CORRUPT BSOD. I have not reinstalled the Windows yet.

 

I've never seen that error code before when doing memory overclocks, and I've gotten a lot of blue screens from unstable overclocks. I'd sat that one of your drivers might've been screwed up by this and a reinstall might be a good idea. 

[Darkbahn]

On 2/7/2023 at 7:32 PM, mrpkv said:

Hello,

 

I have the following configuration:

 

ASUS ROG MAXIMUS HERO Z790

Intel i9-13900K

32GB(2x16) Kingston Fury Renegade at 7200MHz

 

With XMP I and Profile 1 enabled at 7200MHz, MCE disabled, CPU limited at 253W I get application crashes a lot (eg. Steam, Microsoft Office Apps, Chrome STATUS_ACCES_VIOLATION, a lot of Windows Explorer crashes and crashes in games after some time of playing). Also, during the AIDA64 system stability test I get hardware failure after 30 seconds and BSODs with error code MEMORY_MANAGEMENT. I must mention that running Windows Memory Diagnostic or MemTest86 reports 0 errors, I have done these tests multiple times. 
 

When I lower the frequency from 7200MHz to 6933MHz but keep the same timings and voltages I notice improvements, AIDA64 system stability test does not report hardware failure anymore and the crashes and BSODs are less frequent, but they still exist. OCCT memory test also runs fine for 15 minutes at 6933MHz, did not run it for longer. 
 

It also feels like a matter of luck sometimes, as I had days when I saw 0 errors in the reliability monitor, but the next day Windows Explorer or Steam start crashing as soon as I start up the PC and I need to restart otherwise the keep crashing in a loop. 

For a couple of weeks I’ve been trying to figure what is the reason for my system being unstable.

 

I don’t think there is a problem with the CPU or memory, I think there is a setting problem with the way I configure the BIOS or the drivers are problematic as I let ASUS ARMOURY CRATE, NVIDA GeForce Experience and Windows Update install my drivers. 
 

Thank you!

Can you create and share, aida64 reports. Ram xmp profiles screen. 7200mhz also 6800mhz should be there.