ROMED8-2T not booting due to post code 90 (DXE_BDS_STARTED)

Hello,

My system is stopping when booting at Post Code 90 (DXE_BDS_STARTED).
The board has had an RMA done so I know the new board is good.
Nothing else is attached to she system but the CPU and Memory.
Clearing/Upgrading the BIOS/BMC does not resolve the issue.
I am running version 3.7 for BIOS and version 2.02 for BMC.

CPU: AMD EPYC 7313 Milan 16 Cores/32 Threads 3.0GHz Processor 128MB L3 Cache 155W
RAM: HMA84GR7AFR4N-VK HYNIX 32GB (1X32GB) 2RX4 PC4-2666V DDR4 MEMORY

Here are the logs I exported.
It took 3 weeks to get the board replaced so any help would be appreciated.

warning.log
info.log
err.log
dmesg
debug.log
crit.log

Did you check the “Dr. Debug”? The code to meaning is in the manual.

Its leads to DXE_BDS_STARTED which is Boot Device Selection.
Other forum posts that have the same issue don’t have a resolution for it.

Correct me but if you only have the CPU & RAM installed and the DR Debug says you are at the “Boot Device Selection” wouldn’t that mean you need to attach a boot device?

The action Items I would do in your shoes.
#1 Attach a bootable device & set it to be the boot device in the bois. Save/exit.
#2 Reseat the CPU/Memory
#3 Contact Asrock support (again) for the recently RMA’d board.

Good luck and if you do find a resolution please share. The next person will probably appreciate it.

2 Likes

Thanks for the helpful tips.
The main issue is that the system doesn’t boot far enough to get to BIOS or do anything.
Also since they sent me a new board for the RMA I’m not sure what they can do.
I got another CPU to test with and the system is booting properly now.

The broken CPU does look to have contacts that seem “dirtier?” but other than that I’m not sure what could have caused this issue.
I’ve posted about issues with this CPU and board before but the system didn’t get into this state until it just rebooted itself one day.

Is is normal for bad/dirty contact on the mount to kill a CPU?
I’m open to any thoughts or ideas since I’m at a loss about what caused the CPU to fail.

I’m gonna try to fix the broken CPU now that I have a working one, the only solution I have is to clean the contacts.
I’ll add some pictures later of what it looks like before I do anything.

It does look like it is kinda beat up and there is something weird on the pins next to the side.
I’m not sure what the white color is. I must have missed it because you can only see it and other things from certain angles.
Specifically this photo, there are more on imgur.
It looks like the contacts are scratched but I’ve only ever used it in the system.

So the issue is roughly resolved but I thought I would post the steps I took and the timeline.
The TLDR is at the bottom.
About 1-2 months ago the server that had the CPU that I purchased unexpectedly stopped passing POST. After this time I went through the process of getting help from AsRock support staff for the motherboard (ROMED8-2T) that took a week and which ended in a RMA which took 3 weeks to be received but which did not resolve the issue. I tried to RMA the CPU with AMD but they informed me this was a OEM part and to reach out to the OEM as this part is not under warranty. (It is off ebay, I just reached out a short time ago). After this I went and had the CPU inspected under a microscope, and cleaned. There were no external issues that were seen with part but when placed in the motherboard is still does not POST with the last two Post Codes being AC-90. At this time I had no choice but to purchase another CPU to verify that it was the part. This came out to be true.
If anyone has any other ideas I can try before I can’t return the new CPU it would be highly appreciated. I’ve only had the CPU since Feb/Mar of this year and I am confused in how it spontaneously failed. It could be a normal failure but that would be unfortunate.

P.S: If anyone has experience with troubleshooting failed CPU’s I would not mind to send this to you. I would be interested to at least learn what failed and possibly how.
TLDR: Replaced CPU - Motherboard Boots - System Runs

Sorry to hear that. At least you seem to have confirmed its the cpu. Hopefully the seller will take care of you here.
My own experience is not entirely dissimilar.
New/used epyc milan from a well regarded seller on ebay. Bought cpu/mem/mobo in july. Mobo reporting corrected ecc errors in 1 slot near cpu no matter which dimm is there. Spent months dealing with asrock and samsung assuming it a faulty install, mobo or dram. 2 mobo’s and months later i worked with the seller. Troubleshot with him and bought a tyan mobo just to change up the ecosystem. Tyan shut down the dimmslot closest to cpu reporting error. Cpu has to have a defect 100%. DHL has it enroute back to China to the seller who will swap it.
The seller has been good about it so far. Just been a hassle since purchasing because i never assumed it would be the cpu.

Good luck to us both.

This topic was automatically closed 273 days after the last reply. New replies are no longer allowed.