Thank you for responding. Now my question is - I downloaded the v0704 Bios file, but I see no "BIOSRenamer" ? I thought you supposed to run that BIOSRenamer before copying the file to a USB? or we don't need to do that??
There's 2 ways to quickly identify if you have the problem one much quicker than the other. Run blender benchmark with scene victor. It will crash in warmup/loading scene.
The even quicker way is to install XTU run avx 512 for most of the time 5 mins is enough and more often than not it will hardlock in less than 20 seconds.
2x16bg will aggravate this problem more. 2x8gb will alleviate the problem to an extent as ii could prime stable i could aida i could hci memtest I could game but i couldn't XTU avx 512 for 5 minutes or blender victor scene and pass ever. I've held out hopes that it was all bios but as the reports trickled in that people rma'd board or bought multiples then finally 3 boards later there problems magically disappeared and or used another brand and no more problems and since i've verified it was not my memory or just a bad cpu or bad imc in my cpu i'm really just leaning towards there are some faulty boards causing everyone these problems.
Vccio/vccsa alleviated but did not make the problem go away it just bought me maybe 5-10 minutes runtime in XTU sometimes. Setting tfaw to 16 in my testing seems to really magnify the problem and any decent set of samsung Bdie memory can do 16 TFAW just fine so it's not the memory causing this issue.
Higher speed memory also makes the problem worse. I tested 3600 3400 3200 2933 and lower. The only time I ever could pass was defaults loaded.
The same 2x16gb memory/cpu can do 4000 16-16-16 in another brand ITX board HCI stable, Prime stable, aida64 stable, tm5 anta777@extreme stable and XTU stable so that rules out the bad memory or weak IMC possibility.
The vccio/vccsa in all vendors boards is overkill. For like 2x16gb 3600 memory you should really only need 1.15 SA 1.10 IO and then it scales up from there so like 3800 memory 1.20 SA 1.15 IO and 4000 1.25 SA and 1.20 IO you can probably start with similar scaling with 2x8gb but i would try 1.10 SA 1.05 IO for 3600 and scale up from there as 2x8gb should require less.
It's possible those random crashes after having system open for hours etc might be related to sleep/wake from sleep and or leaving cpu to do its thing and power saving features on. You can try shutting all that stuff off and if it no longer crashes it could be related to that.
Also make sure you have the latest update for all the ethernet drivers. Those were causing random hardware errors for some users.
The hardlock in avx xtu and blender victor scene prevented me from even bothering or even worrying about any other potential issues.
Maybe I should try my board again. I did send asus a video of the hardlock so maybe they finally fixxed it. video showed defaults loaded then load xmp boot run xtu crash in 5 secs.
None of the betas or 0607 solved my problem however. I'm fairly certain 0704 I was experiencing the same issue i just have not tested 0707 public.
Here's a video of what I was experiencing and what I sent to ASUS. https://streamable.com/cn51cj