Massive issues with AGESA 1.0.0.5c on TUF B650M

[Paul_Anderegg]

Updated the BIOS which prepares the MB for the new X3D chips which I will be buying when they drop.

First issue was expected, DRAM orange LED nightmare, multiple reboots and forced resetting to get it to even get to the Windows log-in screen, and from there just blue screen after blue screen. Finally got the DRAM to "stick" and got into Windows, ran OCCT to test the RAM stability, it was fine.

Rebooted, blue screen again. Forced restart, now BIOS screen saying new CPU installed and NV TPM is corrupted Y to rest N to continue. I selected no, Windows log-in, instant Blue screen and hang. I did not touch the CPU.

I should also mention that simply trying to scroll through the BIOS locks up the BIOS, requiring reset to get things selectable again, almost every single time. I am reverting if possible, assuming my Windows install is not BRICKED from the BIOS update corrupting my damn TPM.

Just a warning to those considering this update. These issues were DEFAULT BIOS settings as well as when I applied my simple EXPO +200 CPU boost and -30 curve as only modifications. If you are planning an X3D update, perhaps best to sort the new BIOS prior to dropping the new chip in.

Paul

[SkyBeam]

Also have a look at this thread.

I faced similar issues yesterday and the day before trying to upgrade my X670-P to BIOS 1222. I was totally unable to find any valid memory configuration for my 4x G.Skill F5-6000J3038F16G setup at DDR5-6000. It perfectly works out of the box on BIOS 0823.

I even tried manually setting the timings and MC voltage at 1.4V+ with some sort of success but did not get it completely stable. I am usually using memtest86+ for memory validation. The default settings on EXPO-1 in BIOS 0823 are working fine. In BIOS 1222 it won't even POST with default settings. Typically with blinking power button and 1x long, 2x short beep.
So only option here would be to run at DDR5-3600 JEDEC timings.
I did roll back to BIOS 0823.

Well, I also tried curve optimized and also here I am not able to get anything lower than -18 being completely stable on my 7950X - but I guess this is silicon lottery. However I do recommend also for you to either set it to auto for testing purposes or testing less aggressive optimizations. I can also boot Windows 11 with -30 but it's not actually stable in all situations causing crashes, system freezes and spontaneous reboots randomly.

[nite2k]

Adding my two cents here as well. Different form factor and still the same result with the new BIOS that was released on Friday. My board will boot Windows as long as I still with factory defaults or JEDEC RAM speed (4800mhz). The moment I make one change, which is to load the EXPO I profile, the system will not POST).

[chemie]

Seems massive regression with 1222 for memory stability. I reverted after trying 1222. My issue was the "memory context restore" had to be enabled otherwise Windows would insta-crash. With 0821, this is not an issue so gives faster boot. I found no other advantages to 1222 (it did not solve the "crash when wake from sleep if iGPU is disabled" issue I have. I guess I was "lucky" that 6000 did work if the memory context restore was "auto" (forced retrain each boot). This was on B650E-F

[decimus28]

nite2k wrote:
Adding my two cents here as well. Different form factor and still the same result with the new BIOS that was released on Friday. My board will boot Windows as long as I still with factory defaults or JEDEC RAM speed (4800mhz). The moment I make one change, which is to load the EXPO I profile, the system will not POST).

The Crosshair Extreme x670e has the same issue with it's x3d BIOS(0922) for me.