So, I have home lab(I have ZFS, qemu-kvm and kubernetes there), based on Asus x-399-a, tr-1920x, 64GB RAM. And I bought tesla m10 for this home lab. I’ve tested this GPU in the three different PC:
And it works as expected - I want to use this GPU for Ollama/LLM inference
But when I’m trying to use this GPU with my home lab’s mobo: Asus x-399-a it can’t finish the post:
It always in stuck with this non-sense updating RGB
Of course, I’ve disabled everything related to RGB in BIOS. Also, I tried a lot of other things, like upgrading BIOS to the next version, like downgrading video bios of Tesla m10 to previous version.
I wrote on Tom’s hardware what I already tried to do: Question - Boot issue with Tesla M10 ? | Tom's Hardware Forum
I’m considering radical solutions, like custom BIOS ROM for mobo. Or kinda of hardware solutions for miners: is it possible to switch on GPU when my Linux loaded, and just rescan PCI devices?
- very old machine - based on Asus desktop motherboard, with Xeon x5680 (initially it had i7-920). Sorry, don’t remember the exact name of mobo.
- another machine, with MSI motherboard (sorry, don’t remember mobo’s name too) and i9-9900ks CPU
- ASRock X670E PG Lightning + Ryzen 7 7800X3D
And it works as expected - I want to use this GPU for Ollama/LLM inference
But when I’m trying to use this GPU with my home lab’s mobo: Asus x-399-a it can’t finish the post:
It always in stuck with this non-sense updating RGB
Of course, I’ve disabled everything related to RGB in BIOS. Also, I tried a lot of other things, like upgrading BIOS to the next version, like downgrading video bios of Tesla m10 to previous version.
I wrote on Tom’s hardware what I already tried to do: Question - Boot issue with Tesla M10 ? | Tom's Hardware Forum
I’m considering radical solutions, like custom BIOS ROM for mobo. Or kinda of hardware solutions for miners: is it possible to switch on GPU when my Linux loaded, and just rescan PCI devices?
