Ha Dimadima, what makes you think it only takes ECC?
i.e. the specs on the DS1 say only ECC but the DS4 for example can take either.
Hence I'm a bit curious but no spare sticks.

--- Yes I can do but it will be later on, currently got the device in production.

Ah yes, not sure why I had this assumption, I must have confused with the Avoton board which I remember was a bit more restrictive in memory choices.

As per the spec:
Supports 1.2V DDR4 memory
Single & dual rank modules supported
ECC & non-ECC modules supported
RDIMM supported up to 32GB per module
UDIMM supported up to 16GB per module
1600/1866/2133 MHz

I can hardly imagine Datto would have removed the non-ECC support then. Sounds a much bigger change than slapping a 10G nic with an RJ45 PHY in place of SPF+ but who knows now it’s Datto...

I do not have any non-ECC DDR4 to test it...

Wow, already in production? Not going with crunching primes and memtesting for at least a week?;-)

Sorry for the delay guys but here is my working version finally.
@Lost_N_BIOS, here are my MAC Addresses:

E0D55E60F759
E0D55E60F75A

Thanks @Mekw LAN MAC ID's both stored in NVRAM - GUID >> 9669E125-FEDF-43F7-891A-5AF85EFCDEFC (GBTData module, third one, first two do not contain this data in your F11 BIOS)


In @dimadima first dump sent (MY_F06.bin), there is only two of these but the second one looks like the one that should contain MAC ID's in this BIOS, but it's the zero's.
So, initial issue could have been that and we could have fixed by adding MAC ID there, or maybe that never work and it would always be wrong due to how it was being used by the board since it was incorrect model, was ignoring the ID's, probably looking for LAN MAC ID in GbE where it's often stored instead.
Only way to know for sure is for him to reflash that MY_F06.bin BIOS with me editing in MAC ID to that module for a test.

But, anyway, thank you for your dump, this helps me to know to check these NVRAM modules next time around for similar issues with GBT server boards and this can likely only be fixed via dump from user not stock BIOS edited.
Stock BIOS do not contain these modules in NVRAM, so via dump only. I expected this, stock BIOS usually contains 888888888788 or all zero's in place of LAN MAC ID location, and this is either ignored/replaced with the one onboard or mistakenly flashed in if flashed certain ways.
I can't find the 88 one in stock BIOS, so I assume it's all zeros in these BIOS and no way to find that location due to 80K+ results of 12 zeros in a row