Hi guys,
I’m working on running OEM custom vendor Xeons on Asus C621E Sage motherboard.
I have a pair of Xeon Silver 4112 ES (U0 stepping) cpus just for testing purposes, and a pair of 8222CL custom OEM cpus (SSPEC SRFA7, B1 stepping, CPUID 00050657h).
The 4112’s working without any issues with modded BIOS (extracted microcode and CPUID under MMTool + UBU Tools + flashback with USB stick).
The problem arises when I put 8222CL (one or two cpus, it doesn’t matter). Immediately when I push the power button the red LED CATTERR1 shows up, and nothing happens, no Q-code, post etc.
I tried to push 500001C microcode from Asrock BIOS, but it not make any changes (version of this microcode was the same which I saw on the CPU-Z screenshot from the seller,
who used these cpus on Asrock EP2C621D12 WS, see the pictures uploaded below). What I haven’t tested is a swap version of the ME region.
Anyway, in my opinion this problem is not because of the wrong version of microcode/CPUID in BIOS, but something else, like:
1. There is “something” in Intel ME region in UEFI which detects there is a non-standard CPU and forces stop booting with catastrophic error.
2. There is a mismatch in PWM (VRM?) controller firmware, and VRMs can’t handle the right voltage level for these CPUs.
If so, is there any driver in UEFI BIOS to control the VRM controller, or does it need to be flashed directly?
3. This issue is not related with mainboard hardware and only related to Intel ME version (less likely).
I have a bunch of tools and I can make a lot of tests, so if you have any ideas, suggestions, please let me know.
I’ll appreciate that.
If I can manage this issue in the meantime, of course I’ll share this here.
Cheers!
-silencer
