cudaErrorInsufficientDriver cuda config_error ai_generated true

CUDA error: driver version is insufficient for CUDA runtime version (cudaErrorInsufficientDriver)

ID: cuda/cuda-driver-version-unsupported

Also available as: JSON · Markdown · 中文
90%Fix Rate
88%Confidence
1Evidence
2023-08-01First Seen

Version Compatibility

VersionStatusIntroducedDeprecatedNotes
CUDA 12.2 active — — —
NVIDIA Driver 525.60.11 active — — —
Ubuntu 22.04 active — — —
PyTorch 2.1.0 active — — —

Root Cause

The installed NVIDIA driver version is older than the minimum required by the CUDA runtime library loaded by the application.

generic

中文

已安装的 NVIDIA 驱动程序版本低于应用程序加载的 CUDA 运行时库所需的最低版本。

Official Documentation

https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html

Workarounds

  1. 95% success sudo apt-get update && sudo apt-get install nvidia-driver-545 
    sudo apt-get update && sudo apt-get install nvidia-driver-545
  2. 85% success pip install torch==1.13.1 --index-url https://download.pytorch.org/whl/cu117 
    pip install torch==1.13.1 --index-url https://download.pytorch.org/whl/cu117

中文步骤

  1. sudo apt-get update && sudo apt-get install nvidia-driver-545
  2. pip install torch==1.13.1 --index-url https://download.pytorch.org/whl/cu117

Dead Ends

Common approaches that don't work:

  1. 100% fail

    The driver remains at the old version; the runtime still detects incompatibility.

  2. 80% fail

    Environment changes may cause library conflicts but do not fix the fundamental driver-runtime version mismatch.