Hello all,
Just built a new server based on a ROME8D-2T and a 7551p I've spent the weekend trying to get this computer to post with no luck, always shows a "00" code, I've built a number of computers over the years but this is the first server I've built from scratch, I'm hoping you can point out some details I may have missed.
Core Parts List:
- AMD 7551p, The CPU is new came in a sealed box from the local Supermicro distributer
- ROME8D-2T Motherboard, newegg
- 8 x DDR4 3200 RDIMM Kingston KSM32RD8/16HDR
I was originally shopping for a 16 core 7002 and during my hunt i got offered a new 1st gen EPYC 7551p for $500 from one of my employers suppliers (last one). My plan was to start with the 7001 and the ROMED8-2t and later down that track i could upgrade the CPU. I bought memory with this thought in mind so i got 3200 RDIMM's even though the first gen EPYC is only spec'd for 2666. In my previous experience memory generally downclocks with few issues, perhaps i'm mistaken in this case?
The memory was ordered from my local supplier to try and match the QVL list, the only memory i could get in a reasonable time frame was KSM32RD8/16HDR (Hynix) instead of KSM32RD8/16MEI (Micron) that is on the QVL list. my supplier was sourcing based on the motherboard.
I've tried all the usual troubleshooting steps:
- removed all peripherals
- removed all RAM but one stick and cycled my way through the stack based on the manual using C1 for a single stick
- remounted the CPU checked pins
- bios appears to be latest version (possibly a problem for naples?)
- removed motherboard from chassis, disconnect all wires except for ethernet for IPMI and PSU. boot from ipmi
- IPMI logs are clear
I'm going to try to get some slower RAM tomorrow and try to boot with that to see if i can rule in/out the RAM, will also try another power supply.
After a bunch of reading of similar issues i'm leaning toward a motherboard problem just becuse of the lack of a proper "Dr. Debug" code since "00" is not on the list, but since i don't have access to another EPYC cpu it could be a problem trying to 100% figure out which part(s) are causing issues.
Appreciate any insights
Just built a new server based on a ROME8D-2T and a 7551p I've spent the weekend trying to get this computer to post with no luck, always shows a "00" code, I've built a number of computers over the years but this is the first server I've built from scratch, I'm hoping you can point out some details I may have missed.
Core Parts List:
- AMD 7551p, The CPU is new came in a sealed box from the local Supermicro distributer
- ROME8D-2T Motherboard, newegg
- 8 x DDR4 3200 RDIMM Kingston KSM32RD8/16HDR
I was originally shopping for a 16 core 7002 and during my hunt i got offered a new 1st gen EPYC 7551p for $500 from one of my employers suppliers (last one). My plan was to start with the 7001 and the ROMED8-2t and later down that track i could upgrade the CPU. I bought memory with this thought in mind so i got 3200 RDIMM's even though the first gen EPYC is only spec'd for 2666. In my previous experience memory generally downclocks with few issues, perhaps i'm mistaken in this case?
The memory was ordered from my local supplier to try and match the QVL list, the only memory i could get in a reasonable time frame was KSM32RD8/16HDR (Hynix) instead of KSM32RD8/16MEI (Micron) that is on the QVL list. my supplier was sourcing based on the motherboard.
I've tried all the usual troubleshooting steps:
- removed all peripherals
- removed all RAM but one stick and cycled my way through the stack based on the manual using C1 for a single stick
- remounted the CPU checked pins
- bios appears to be latest version (possibly a problem for naples?)
- removed motherboard from chassis, disconnect all wires except for ethernet for IPMI and PSU. boot from ipmi
- IPMI logs are clear
I'm going to try to get some slower RAM tomorrow and try to boot with that to see if i can rule in/out the RAM, will also try another power supply.
After a bunch of reading of similar issues i'm leaning toward a motherboard problem just becuse of the lack of a proper "Dr. Debug" code since "00" is not on the list, but since i don't have access to another EPYC cpu it could be a problem trying to 100% figure out which part(s) are causing issues.
Appreciate any insights