Score:0

Initialization 91 stuck of Supermicro OEM (HPE) GPU server

cn flag

I just set up HPE SX40 (Supermicro SYS-1029GQ-TVRT).

The system stuck on boot with "initializing ... 91". I tried some crude test like remove all memory, and got relevant error message.

So, 1)removed all pcie cards (including NVIDIA SXM2 adapter boards) and boot -> no luck

2)Followed FAQ to reset CMOS (https://www.supermicro.com/support/faqs/faq.cfm?faq=18922) -> no luck

3)Direct bios update (https://www.supermicro.com/support/faqs/faq.cfm?faq=20491) -> no luck at all.

Any comments in Supermicro stuck on 91 also didn't work.

Does anybody knows what 91 exactly means for X11DGQ (the board built by supermicro)?

I am suspicious about the number of gpus (I have only one V100 gpu, so connected in SXM2 slot 3 following HP configuration of two cards (slot 3 and slot 2, for cpu1 and cpu2), but it still make no sense about 1), since same 91 error shows up after removing all pcie connections.

Score:0
cn flag

All above trials are no-luck, but after struggling for three days, I decided to disassemble all and clean-reassemble.

It boot!!! I figured out something was wrong in cpu2. cpu issue also might cause 91! since it's pcie lane related : )

mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.