afazekas

Hi, In the referenced container version it can work with a gfx1030 system. docker.io/rocm/dev-ubuntu-22.04:5.6-complete apt show rocm-libs -a ``` Package: rocm-libs Version: 5.6.0.50600-67~22.04 Priority: optional Section: devel ``` But in...

Thanks for the quick replay. I am using the system defaults so `runc `. The different on the strace, when used kubectl/oc. $ strace -f -p -v -s 4096 -tt...

log has something strange which sounds related. Probably not just from the bad cases. ``` Aug 16 09:24:52 master-0.ocp.openstack.lab crio[3710914]: time="2024-08-16 09:24:52.108278647Z" level=debug msg="Response: &ExecSyncResponse{Stdout:[FILTERED],Stderr:[],ExitCode:0,}" file="otel-collector/interceptors.go:74" id=2e92020e-4> Aug 16 09:24:52...

Can you try in a "CRC VM for OpenShift 4.15.17", or can you suggest a crc version when it is not reproducible ?

RTX 6900 works for me (main branch). I do not see ngl passed in the podman invocation from your output: `exec_cmd: podman run --rm -i --label ai.ramalama --name qwen --env=HOME=/tmp...

Since ramalama tring to use the ramalma image instead of the rocm image likely it misses something about the GPU. I do not remember anything special about f41and rocm it...

I wonder what these commands do in your system: echo /sys/bus/pci/devices/*/mem_info_vram_total cat /sys/bus/pci/devices/*/mem_info_vram_total A number above 1073741824 should enable rocm usage by default. --ngl might not be passed when no...

Looks like when HSA_OVERRIDE_GFX_VERSION is specified the gpu detection is not running so the HIP_VISIBLE_DEVICES=0 is not set.

You can set the HIP_VISIBLE_DEVICES=0 when you set HSA_OVERRIDE_GFX_VERSION. Can you check is it solves everything even the automatic image selection ? The code consider 'HSA_VISIBLE_DEVICES' also gpu env, which...

IMHO,not just basic avx needed. In case your NUC supports thunderbolt or other type-c with PCIe tunneling (some of the USB4 ports), you can consider an eGPU adapter.