J’ai changé la carte NVIDIA car deux cartes NVIDIA avec 8 Go chacune, elles sont vues par la VM qui est lancé par proxmox :

# nvidia-smi --list-gpus
GPU 0: Quadro M5000 (UUID: GPU-)
GPU 1: Quadro M4000 (UUID: GPU-)
# nvidia-smi      
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 570.86.15              Driver Version: 570.86.15      CUDA Version: 12.8     |
|-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  Quadro M5000                   Off |   00000000:00:10.0 Off |                  Off |
| 38%   37C    P8             13W /  150W |       5MiB /   8192MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
|   1  Quadro M4000                   Off |   00000000:00:11.0 Off |                  N/A |
| 46%   39C    P8             13W /  120W |       5MiB /   8192MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
                                                                                         
+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI              PID   Type   Process name                        GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|  No running processes found                                                             |
+-----------------------------------------------------------------------------------------+

les résultats du test sont les suivants :

# llm_benchmark run
-------Linux----------
{'id': '0', 'name': 'Quadro M5000', 'driver': '570.86.15', 'gpu_memory_total': '8192.0 MB', 'gpu_memory_free': '8110.0 MB', 
'gpu_memory_used': '5.0 MB', 'gpu_load': '0.0%', 'gpu_temperature': '36.0°C'}
{'id': '1', 'name': 'Quadro M4000', 'driver': '570.86.15', 'gpu_memory_total': '8192.0 MB', 'gpu_memory_free': '8110.0 MB', 
'gpu_memory_used': '5.0 MB', 'gpu_load': '0.0%', 'gpu_temperature': '38.0°C'}
At least two GPU cards
Total memory size : 61.36 GB
cpu_info: Intel(R) Xeon(R) CPU E5-2450 v2 @ 2.50GHz
gpu_info: Quadro M5000
Quadro M4000
os_version: Ubuntu 22.04.5 LTS
ollama_version: 0.5.7
----------
....
-------Linux----------
{'id': '0', 'name': 'Quadro M5000', 'driver': '570.86.15', 'gpu_memory_total': '8192.0 MB', 'gpu_memory_free': '3277.0 MB', 
'gpu_memory_used': '4838.0 MB', 'gpu_load': '0.0%', 'gpu_temperature': '65.0°C'}
{'id': '1', 'name': 'Quadro M4000', 'driver': '570.86.15', 'gpu_memory_total': '8192.0 MB', 'gpu_memory_free': '2348.0 MB', 
'gpu_memory_used': '5767.0 MB', 'gpu_load': '0.0%', 'gpu_temperature': '76.0°C'}
At least two GPU cards
{
    "mistral:7b": "16.56",
    "llama3.1:8b": "15.71",
    "phi4:14b": "8.01",
    "qwen2:7b": "15.27",
    "gemma2:9b": "15.81",
    "llava:7b": "17.82",
    "llava:13b": "13.14",
    "uuid": "1a60faf0-e97b-5d47-8de5-03d3b22dfbbc",
    "ollama_version": "0.5.7"
}

Actuellement j’utilise « llama3.1:8b », je suis donc passé le 1.12 (unitilisable) à 15,71 . L’idéal est d’avoir dans les plus de 32 … donc il va falloir trouver deux nouvelles cartes.

Misère.