Asus Z690 TUF D4 Rebooting on lighter load

[postem]

Hi, i came across an issue, that i found some people reporting also on reddit. I was on Bios 707 now im in BIOS 807, problem still happening.
Someone at overclock.net reported the same issue on Strix z690 d4, and the workaround was suggested to keep vcore fixed.

Out of nowhere, some games trigger a hard reboot. I was playing Transport fever 2, and consistently the game was making the computer to reboot, much like a reset. Same thing happened yesterday on Wolfenstein 2, during video playback before a new level. On both situations, the load was minimal, i kept hwinfo logging, and everything, temperatures, current, watts, everything well below 350w total. Windows system log only reports kernel power loss.

Current System:
i7 12700K (stock) + Z690 tuf d4 + 3600 32gb ballistix (QVL) + 2080ti (no oc) + Seasonic gold 850 PSU. Windows 11 Enterprise
Previous system:
i9 9900KS overclocked 5.1 ghz + Asus Hero x + 16gb 3600 + 2080ti (no oc) + Seasonic gold 850w PSU Windows 11 Enterprise

No OC applied, only XMP profile, and this memory is on the QVL. Both with seasonic 850w gold, both on windows 11

Windows, applications, all stress tests pass without issues, including furmark, realbench, aida multi test, memtest64, P95 AVX, Cinebench, i throw every kind of heavy load and the system never trips, only on some games.

On transport fever 2 it was quite remarkable, because once i loaded the game and pan the map the reboot happened. Then i reinstalled the game and NVIDIA drivers, and the problem seen to be resolved, but i still had one reboot on this game.

I was using the same PSU on the previous board+cpu, which i never had any issues, which i would target as probably first culprit in my previous system without issues. Stressed beyond 500W and no issues, not a single failure, reboot or bsod.

Wolfenstein which i get several reboots, reboots me in the lighest of the loads, when im loading a save or its running a video intro of a level. Someone suggested me that i keep vcore fixed, im trying this now, 1.3v.

There is some kind of know BIOS issue that is due to be fixed?

[Silent_Scone]

Hello, if the issue doesn't occur without the XMP profile enabled (Optimised defaults) then the issue is the memory overclock is unstable. XMP not being overclocking is a common misconception.

Check this with a memory stress test such as Karhu RAM Test or MemtestPro (HCI)

[postem]

Silent Scone@ROG wrote:
Hello, if the issue doesn't occur without the XMP profile enabled (Optimised defaults) then the issue is the memory overclock is unstable. XMP not being overclocking is a common misconception.

Check this with a memory stress test such as Karhu RAM Test or MemtestPro (HCI)

Hi thanks for the reply. I already tested several passes of memtest64, gonna do some testing with your recommended programs, but i used a lot of testing and stress tools, never to find a hard reboot. I left for a night realbench and not any issues.

I set fixed vcore to 1.3v, llc4, and so far at least no reboots yet, its difficult to provoke the issue in a consistent way.

One thing that amazed me is that under fixed vcore the request vid seen to be more sane (i know vid is useless in fixed vcore), but requested voltages are much more sane, often up to 0.1v of difference from the stock settings.

[postem]

Silent Scone@ROG wrote:
Hello, if the issue doesn't occur without the XMP profile enabled (Optimised defaults) then the issue is the memory overclock is unstable. XMP not being overclocking is a common misconception.

Check this with a memory stress test such as Karhu RAM Test or MemtestPro (HCI)

Hi tried again, everything on auto including memory, not using XMP, and yet, wolfenstein on load save that is a simply load just reboots the computer.
Im running out of ideas what component could be the issue, if the memory, PSU or the memory or the cpu itself.

The only event i get on windows log is:
The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.

[Kikoarl]

postem wrote:
Hi tried again, everything on auto including memory, not using XMP, and yet, wolfenstein on load save that is a simply load just reboots the computer.
Im running out of ideas what component could be the issue, if the memory, PSU or the memory or the cpu itself.

The only event i get on windows log is:
The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.

I have similar issues with my D4 board. It seems that, even though my QVL RAM is 3200Mhz (which should be SUPPORTED NATIVELY by Alder Lake) it's not supported by the complete dumpster fire the BIOS is. Still running a NOVEMBER BIOS because ASUS seems to have abandoned my board. PRIME Z690-P D4. Can't even run DEFAULT SPEEDS with my RAM without software crashing.

[postem]

Kikoarl wrote:
I have similar issues with my D4 board. It seems that, even though my QVL RAM is 3200Mhz (which should be SUPPORTED NATIVELY by Alder Lake) it's not supported by the complete dumpster fire the BIOS is. Still running a NOVEMBER BIOS because ASUS seems to have abandoned my board. PRIME Z690-P D4. Can't even run DEFAULT SPEEDS with my RAM without software crashing.

Did you get random reboots too? I tried with default memory speeds and it still give me issues.

Im getting NUTS with this issue. I already tried everything, the only other option i have and i will try tomorrow is to change the 32gb kit for my 16gb gskill kit.
Even without having any issues with this ram at least on testing, its a cheaper kit, who knows.

I was kinda not worried because wolfenstein 2 is a bug fest, but today i had another reboot using civilization 6. I know its also i game i had issues in past, but usually the error just means crash to desktop and not a reboot. Battlefield 1, i played 4 hours straight and 0 issues. Frankly i cant say anymore if these issues are from the cpu or the motherboard or memory;

Ii managed to consistently reproduce the issue on wolfenstein 2, i need to keep dying and reloading the save, until it reboots. Since its a very specific circunstance that i can manually trigger that doesnt point to a PSU issue, especially when its a very low load.

Since every time is related it seen to a disk load (either auto saving on civilization or auto loading on wolfenstein) i will probably also test on my second windows that is installed on another SSD.

[Kikoarl]

postem wrote:
Did you get random reboots too? I tried with default memory speeds and it still give me issues.

Im getting NUTS with this issue. I already tried everything, the only other option i have and i will try tomorrow is to change the 32gb kit for my 16gb gskill kit.
Even without having any issues with this ram at least on testing, its a cheaper kit, who knows.

I was kinda not worried because wolfenstein 2 is a bug fest, but today i had another reboot using civilization 6. I know its also i game i had issues in past, but usually the error just means crash to desktop and not a reboot. Battlefield 1, i played 4 hours straight and 0 issues. Frankly i cant say anymore if these issues are from the cpu or the motherboard or memory;

Ii managed to consistently reproduce the issue on wolfenstein 2, i need to keep dying and reloading the save, until it reboots. Since its a very specific circunstance that i can manually trigger that doesnt point to a PSU issue, especially when its a very low load.

Since every time is related it seen to a disk load (either auto saving on civilization or auto loading on wolfenstein) i will probably also test on my second windows that is installed on another SSD.

I don't get random reboots. For me, when I start a memory intensive program such as DREAM3D, it closes without a single error. Nothing to be seen on the event viewer.

However, a friend suggested me to try to install a pirated game from one repacker who is known to use a lot of RAM in the process. It seems that during the install process, the game gets loaded from the downloaded files to the RAM, then after a few seconds, it gets offloaded from the RAM into the installation folder I gave to the installer. This happens a few times during the whole install.

During the first offload process, the installer crashes and gives me two errors, telling me that it's impossible to write on that hard disk and the files are corrupted, so it can't be installed.

My disks are fine. My friend even gave me his external hard disk with a few games so I can try them. He can install all of them while I can't install a single game.

He has a Ryzen 5600x with non QVL RAM.

[Falkentyne]

Try lowering the VRM Switching frequency in Digi+VRM.

[Raad11]

Make sure VRM switching frequency, cpu duty control and phase control are all set to Auto

[postem]

Hi for everyone who replied the thread: problem fixed with a PSU replacement.

My older PSU was 6 years old, but still, i really think the problem was all due when i connected the second EPS cable, and it probably ended in the same internal rail as the gpu and probably during a spike it was causing OCP tripping. Since the PSU was already old and cant really dismiss worn out at least a bit of capacitors, i purchased a XPG Core reactor 850W, the only decent PSU with reasonable price / avaialbility in brazil. This psu also have a single rail.

With the new PSU the tripping stopped completely.