• Members 13 posts
    2026年5月15日 11:15

    一、软硬件信息
    1.服务器厂家:浪潮信息
    2.沐曦GPU型号:单张沐曦曦思N260
    3.操作系统内核版本:4.19.90-89.11.v2401.ky10.x86_64
    4.是否开启CPU虚拟化:是
    5.mx-smi回显:
    mx-smi version: 2.3.1

    =================== MetaX System Management Interface Log ===================
    Timestamp : Fri May 15 11:06:59 2026

    Attached GPUs : 1
    +---------------------------------------------------------------------------------+
    | MX-SMI 2.3.1 Kernel Mode Driver Version: 3.8.23 |
    | MACA Version: unknown BIOS Version: 1.31.1.0 |
    |------------------+-----------------+---------------------+----------------------|
    | Board Name | GPU Persist-M | Bus-id | GPU-Util sGPU-M |
    | Pwr:Usage/Cap | Temp Perf | Memory-Usage | GPU-State |
    |==================+=================+=====================+======================|
    | 0 MetaX N260 | 0 Off | 0000:c1:00.0 | 0% Enabled |
    | 58W / 225W | 56C P9 | 58711/65536 MiB | Available |
    +------------------+-----------------+---------------------+----------------------+

    +---------------------------------------------------------------------------------+
    | Sliced GPU |
    |------------------------------------+---------------------+----------------------|
    | Minor GPU sGPU-Id Compute | Vram Quota | sGPU-Util |
    |====================================+=====================+======================|
    | 000 0 0 80% | 47068/49152 MiB | 0% |
    +------------------------------------+---------------------+----------------------+
    | 001 0 1 20% | 10976/12288 MiB | 0% |
    +------------------------------------+---------------------+----------------------+

    +---------------------------------------------------------------------------------+
    | Process: |
    | GPU PID Process Name GPU Memory |
    | Usage(MiB) |
    |=================================================================================|
    | 0-s0 1797542 VLLM::EngineCor 47068 |
    | 0-s1 1800252 VLLM::EngineCor 10976 |
    +---------------------------------------------------------------------------------+

    二、问题现象
    由于只有一张卡且需要同时运行qwen3和qwen3-embedding模型,之前使用sgpu将单卡分为了48G/80%和12G/20%使用,但是发现设备一旦重启,所划分的sGPU就会不见,显卡sGPU重新变为disabled,求问如何能够让sGPU开启持久化?重启之后的mx-smi回显如下所示:
    mx-smi version: 2.3.1

    =================== MetaX System Management Interface Log ===================
    Timestamp : Fri May 15 11:10:17 2026

    Attached GPUs : 1
    +---------------------------------------------------------------------------------+
    | MX-SMI 2.3.1 Kernel Mode Driver Version: 3.8.23 |
    | MACA Version: unknown BIOS Version: 1.31.1.0 |
    |------------------+-----------------+---------------------+----------------------|
    | Board Name | GPU Persist-M | Bus-id | GPU-Util sGPU-M |
    | Pwr:Usage/Cap | Temp Perf | Memory-Usage | GPU-State |
    |==================+=================+=====================+======================|
    | 0 MetaX N260 | 0 Off | 0000:c1:00.0 | 0% Disabled |
    | 35W / 225W | 50C P0 | 666/65536 MiB | Available |
    +------------------+-----------------+---------------------+----------------------+

    +---------------------------------------------------------------------------------+
    | Process: |
    | GPU PID Process Name GPU Memory |
    | Usage(MiB) |
    |=================================================================================|
    | no process found |
    +---------------------------------------------------------------------------------+

  • arrow_forward

    Thread has been moved from 产品&运维.

  • Members 458 posts
    2026年5月15日 14:03

    尊敬的开发者您好,SGPU不支持持久化,可以通过mx-smi开机自动配置