Crosshair VIII Hero (WiFi) what's going on with new BIOS versions ?

[GoLLuM13]

Hi everyone,

Right now I'm on the 3801 version and everything is stable and works perfectly but as soon as I try a higher version 3904 or 4006 for example, I have multiple and random issues whether at boot with a black screen and different Q-Codes each time amongst them 02 - 22 - FA - F9, I also had error messages such as "BIOS Recovery mode has been detected" or "Invalid signature detected" (for those two I guess it's because of the USB drive I used to flash the BIOS), when I manage to load Windows 11 and reboot whether I have the same symptoms or I have a BSOD while loading Windows or after it, and all of these problems happen mostly when I set up the RAM whether manually (as soon as I reach 2800Mhz) or via D.O.C.P., and they disappear as soon as I set memory to "auto" or flash back 3801 BIOS version

Am I the only one having these kinds of issues ? What am I doing wrong ? Or is it like I think a problem with the new BIOSes versions since everything is working perfectly with older versions 3801 and previous ?

Of course, I flashed again and again multiple times, with different USB drives, from different downloads for a same version with Clear CMOS before and after flashing, tried with Bios Flashback and EZ Flash 3

My build : Crosshair VIII Hero (WiFi) - Ryzen 9 3900x - Corsair Vengeance RGB Pro CMW32GX4M4C3000C15 (8GB*4 - 3000Mhz - C15) - Sabrent Rocket 4.0 500G - nVidia RTX 3090 FE
OS : Windows 11 Pro x64 (up to date)

[Escorpiom]

Those sticks appear to be Samsung C-die, and you have 4 of them.
It's a known issue with X570 Asus boards and 4 modules Samsung C-die.
You can try to stabilize them by reducing the voltage (1.33v-1.34v, more info is available on the net.

Cheers.

[GoLLuM13]

Escorpiom wrote:
Those sticks appear to be Samsung C-die, and you have 4 of them.
It's a known issue with X570 Asus boards and 4 modules Samsung C-die.
You can try to stabilize them by reducing the voltage (1.33v-1.34v, more info is available on the net.

Cheers.

Hi Escorpiom, and thank you for your answer, according to CPU-Z I have single ranked SK Hynix modules, (and if I remember correctly when I used to use Thaiphoon on Windows 10 it was CJR-Dies) and they worked like a charm since day one with this motherboard (late 2019 - early 2020) until version 3801 included but starting from 3904 problems started. Nonetheless, I'll try to lower voltage from 1.35 to 1.34 or 1.33 thanks again

[GoLLuM13]

I tried on 4006 to reduce the voltage from 1.35 to 1.34 and then to 1.33 same symptoms after reboot BSODs and "BIOS Recovery mode has been detected" and also some black screens with different Q-Codes, and like before everytime I let everything RAM related to auto it "works" again (but it's set up to 2114Mhz CL22) and of course, as soon as I got back to 3801 and miracle everything is working "perfectly" again, no more BSODs, black screens or warnings.
That's why my conclusion tends to be that there's something "wrong" with the code of these newer versions, but if I'm missing a point I'm open to try

EDIT 1 :
I just saw on another thread right here different build (with only 2 memory sticks) only common points are :
Having an Asus motherboard - Zen 2 CPU (Ryzen 3000) - and going from a BIOS based on AGESA V2 PI 1.2.0.3 Patch C to a BIOS based on AGESA V2 PI 1.2.0.6b
And this is confirming what I was thinking about something wrong in the newer BIOSes

EDIT 2 :
I checked once again Memory QVL of my motherboard with Ryzen 3000 Desktop CPUs and I've found my kit in it, confirming what I was saying earlier, it's a SK Hynix single ranked kit supposed to work with either 1, 2 or 4 sticks at 1.35v

EDIT 3 :
Same behaviour described in this post and this one too, in the first one it has been "fixed" with a CPU swap from Ryzen 3000 to Ryzen 5000 :confused:
Asus (and/or AMD) has to do something about this

EDIT 4 :
I checked in another brand's forum, they have same issues with BIOSes that are based on a higher version than AGESA V2 PI 1.2.0.3 Patch C

[Escorpiom]

If your memory is indeed Hynix, yes than you are right and my findings do not apply to the issue you are having.
Corsair has certain versions of those modules with Samsung C-die, you can look at it. If it is version 4.32, it is Samsung C-die.

About the differences in UEFI - Yes, I had to change my memory voltage also with the last 4004 UEFI.
So Asus (or AMD by means of a new AGESA) has changed something that affected memory.

In my case, I do have 4 modules C-die 3200Mhz and they won't run anything higher than 3000Mhz.
This has been an issue since I bought the setup.

Cheers.

[GoLLuM13]

Thanks again for your help Escorpiom, now I'm pretty sure the problem is in the AGESA beginning from 3904 and not in my memory kit or some wrong setup.
In 3801 version and earlier I could even reach 3200 Mhz effortlessly or even 3600 Mhz with some tweaking (for tests), while in 3904 and 4006 I can't even run them correctly at 2800 Mhz, and since other brands have the same behavior with the newer AGESA, so I guess that's it

Have a nice day

[LeftEyeNine]

Stay exactly where you are, my friend, as you are not the only one having weird issues with the latest AGESA versions.

While I was on 4004, my newest SSD (3 months old) disappeared completely. Reverting back to 3801 was the only solution.

[GoLLuM13]

@LeftEyeNine I also had this issue after some random BSODs, but I managed to find a "workaround" to make it appear again by doing a Clear CMOS but yes I will clearly stay on 3801 until the next "stable" BIOS (at least I hope there will be one), thanks for your answer it's reassuring me

[karmal]

Same problem as mine with bios 3801 all stable I have 4 modules of 32 gb for a total of 128 and they work at 3600 mhz (HyperX FURY Black HX436C18FB3K4 certified on the asus list) on cpu 3950 x with cross hero Viii wifi very occasionally I only have one problem random the pc starts but the screen remains black I have to physically turn off with the key and turn it back on and everything starts but I don't know if it is due to bios ra mo what ... this problem happens randomly I can start up 100 times and nothing happens ... instead with the following bios a disaster that is not stable at all

[GoLLuM13]

karmal wrote:
Same problem as mine with bios 3801 all stable I have 4 modules of 32 gb for a total of 128 and they work at 3600 mhz (HyperX FURY Black HX436C18FB3K4 certified on the asus list) on cpu 3950 x with cross hero Viii wifi very occasionally I only have one problem random the pc starts but the screen remains black I have to physically turn off with the key and turn it back on and everything starts but I don't know if it is due to bios ra mo what ... this problem happens randomly I can start up 100 times and nothing happens ... instead with the following bios a disaster that is not stable at all

Thanks for the feedback, There's something wrong in the newer AGESA versions (1.2.0.5 and 1.2.0.6) that kinda reassuring to see it's not our motherboard or any other component that is faulty but the BIOS Version is (kinda)