"PEI Intel MRC Execution" on SuperMicro A2SDi-8C-HLN4F

diskdiddler

Wizard
Joined
Jul 9, 2014
Messages
2,377
I wanted to upgrade my 5 year old NAS with 2 more sticks of RAM, so I was very pedantic and googled my butt off BEFORE ordering new memory.

I see this issue here:
and here:

(Note I've tested with ONLY the new RAM in the 2 known working slots, same error)

Old memory, top, new memory bottom:
KB8r7Vp.jpeg
yoTPNR8.jpeg



It's VERY similar ram, VERY! :(
I'm pretty frustrated, I know about RAM pickyness, I thought my pedantry would have this one in the bag :mad:

Here's my board:

Here's the original RAM I purchased in 2018:
"Supermicro compatible MEM-DR416L-HL03-ER24 16GB (1x16GB) DDR4 2400 (PC4 19200) 2Rx8 288-Pin 1.2V ECC Registered Memory
NEMIX RAM is manufactured using top brand chips such as Micron, Samsung, Hynix or Elpida and meets or exceeds manufacturer specification along with JEDEC standards."


Here's the new RAM I just got this year on Amazon:
"NEMIX RAM Supermicro Compatible MEM-DR416L-HL03-ER24 16GB DDR4 2400 RDIMM"
About this item
Supermicro Certified MEM-DR416L-HL03-ER24
16GB (1x16GB) DDR4 2400 (PC4 19200)
ECC Registered RDIMM Server Memory RAM
Lifetime Replacement Warranty


Identical part#!


I would assume that NEMIX ram are selling a part which is extremely close to the real thing but actually incompatible?
Who is at fault here, SuperMicro aren't going to release a new firmware to fix me up, feels like NEMIX have 'adjusted their recipe' ?
As an Aussie, I ordered this to my mother in law in the US 8 weeks ago, takes time for me to visit and pick up.

Do I badger Amazon, or Nemix here?
Based *purely on the very very clear descriptions* I feel like I'm not in the wrong here?
 
Last edited:

diskdiddler

Wizard
Joined
Jul 9, 2014
Messages
2,377
Ok I just want to update the post to help others.


I updated my firmware to 1.9b from 1.8 and when I did this, I noticed even with my just 2 good sticks of memory, the system took a long time at the "PEI Intel MRC Execution" stage.

I decided to open the system up, pop the memory back in that's "bad" and left it, for a long time. I think it's possible it's 'training' the RAM as eventually it's come up fine with no issues.

I have encountered, what I think, is a small bug entirely unrelated, but this issue is fixed
 
Top