Conda Linux install does not find supported CPU

[MWR1]

Hi.  Any help/guidance would be greatly appreciated.

Following the directions here:
https://docs.openvino.ai/2025/get-started/install-openvino/install-openvino-conda.html 

OpenVINO appears to be installed with no errors during install process:

# python -c "from openvino import get_version;print(get_version())"
2025.2.0-51-4967d4f6d70

However, checking for available devices yields nothing:

# python -c "from openvino import Core; print(Core().available_devices)"
[]

I expect 'CPU' to be listed. My CPU is Xeon gen 4 (SPR) which is on the supported list.

# lscpu
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Address sizes: 46 bits physical, 57 bits virtual
Byte Order: Little Endian
CPU(s): 32
On-line CPU(s) list: 0,1,3,4,6,7,9-11,14,16-22,24,26,31
Off-line CPU(s) list: 2,5,8,12,13,15,23,25,27-30
Vendor ID: GenuineIntel
BIOS Vendor ID: Intel(R) Corporation
Model name: Intel(R) Xeon(R) w5-3435X
BIOS Model name: Intel(R) Xeon(R) w5-3435X
CPU family: 6
Model: 143
Thread(s) per core: 2
Core(s) per socket: 16
Socket(s): 1
Stepping: 8

I believe I am meeting the system requirements listed here: System Requirements — OpenVINO™ documentationCopy to clipboardCopy to clipboard — Version(2025)

OS: Ubuntu 22.04.5 LTS
Kernel: 6.8.12-11-pve
Python: 3.12.2

Any ideas why this might not be working?

Thanks.

[Zulkifli_Intel]

Hi MWR1,

Thank you for contacting us.

I tried on my side and was able to detect the CPU (not that I'm not using an Xeon processor).

Can you run the Hello Query Device sample to check if the device is available?

Regards,

Zul

[MWR1]

No devices found:

(openvino) root@OpenVINO:~# python hello_query_device.py 
[ INFO ] Available devices:

[Zulkifli_Intel]

Hi MWR1,

Thank you for sharing the output. Can you uninstall and reinstall OpenVINO? Or try installing the OpenVINO base package and see if the device can be detected.

Regards,

Zul

[MWR1]

I created a new Conda env and reinstalled OpenVINO, but same result.  I also tried to use the PIP installation instructions, but same result.  I'm only using the base package in these install attempts. 

(base) root@OpenVINO:~# python -m venv openvino_env
(base) root@OpenVINO:~# source openvino_env/bin/activate
(openvino_env) (base) root@OpenVINO:~# python -m pip install --upgrade pip
Requirement already satisfied: pip in ./openvino_env/lib/python3.13/site-packages (25.1.1)
(openvino_env) (base) root@OpenVINO:~# pip install openvino==2025.2.0
Collecting openvino==2025.2.0
  Downloading openvino-2025.2.0-19140-cp313-cp313-manylinux2014_x86_64.whl.metadata (12 kB)
Collecting numpy<2.3.0,>=1.16.6 (from openvino==2025.2.0)
  Downloading numpy-2.2.6-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl.metadata (62 kB)
Collecting openvino-telemetry>=2023.2.1 (from openvino==2025.2.0)
  Downloading openvino_telemetry-2025.2.0-py3-none-any.whl.metadata (2.3 kB)
Collecting packaging (from openvino==2025.2.0)
  Using cached packaging-25.0-py3-none-any.whl.metadata (3.3 kB)
Downloading openvino-2025.2.0-19140-cp313-cp313-manylinux2014_x86_64.whl (47.6 MB)
   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 47.6/47.6 MB 10.3 MB/s eta 0:00:00
Downloading numpy-2.2.6-cp313-cp313-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (16.5 MB)
   ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 16.5/16.5 MB 83.6 MB/s eta 0:00:00
Downloading openvino_telemetry-2025.2.0-py3-none-any.whl (25 kB)
Using cached packaging-25.0-py3-none-any.whl (66 kB)
Installing collected packages: openvino-telemetry, packaging, numpy, openvino
Successfully installed numpy-2.2.6 openvino-2025.2.0 openvino-telemetry-2025.2.0 packaging-25.0
(openvino_env) (base) root@OpenVINO:~# python hello_query_device.py 
[ INFO ] Available devices:

I'm not getting any errors or warnings, so I suspect there is something about my configuration that OpenVINO doesn't like, but as far as I can tell, this should be a supported config.  I tried installing in a python 3.12 env and 3.10.  I also tried OpenVINO 2025.1.0.

[MWR1]

I found that on the same hardware and bare metal OS (Proxmox) I can get OpenVINO to see both the CPU (and GPU) in a VM container.  I am trying to run OpenVINO in a privileged LXC however and not a VM.  Comparing the non-working LXC to the working VM:

	Working VM	Not working LXC
Linux kernel	6.14.0-23-generic	6.8.12-11-pve
OS	PRETTY_NAME="Ubuntu 25.04"	PRETTY_NAME="Ubuntu 22.04.5 LTS"
CPU	Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Address sizes: 46 bits physical, 57 bits virtual Byte Order: Little Endian CPU(s): 20 On-line CPU(s) list: 0-19 Vendor ID: GenuineIntel Model name: Intel(R) Xeon(R) w5-3435X CPU family: 6 Model: 143 Thread(s) per core: 1 Core(s) per socket: 20 Socket(s): 1 Stepping: 8	Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Address sizes: 46 bits physical, 57 bits virtual Byte Order: Little Endian CPU(s): 32 On-line CPU(s) list: 0,1,3,4,6,7,9-11,14,16-22,24,26,31 Off-line CPU(s) list: 2,5,8,12,13,15,23,25,27-30 Vendor ID: GenuineIntel BIOS Vendor ID: Intel(R) Corporation Model name: Intel(R) Xeon(R) w5-3435X BIOS Model name: Intel(R) Xeon(R) w5-3435X CPU family: 6 Model: 143 Thread(s) per core: 2 Core(s) per socket: 16 Socket(s): 1 Stepping: 8

Comparing the differences, I believe both OS/kernel combos are on the supported list.  Only thing I see is the presence of "Off-line CPU(s)" in the lscpu report out.  Might that have something to do with OpenVINO's device detection algorithm?

[Zulkifli_Intel]

Hi MWR1,

OpenVINO does not officially support LXC, and it requires additional configuration to get it working. If you want to use a container, Docker is the best option.

Regards,

Zul

[Zulkifli_Intel]

Thank you for your question. If you need any additional information from Intel, please submit a new question as this thread is no longer being monitored.