wsl配置cuda

最近想用gpu并行开发一些算法，开发程序在linux下比较方便，于是打算折腾下wsl下的cuda配置。

step1. 查看nvidia配置信息

在命令行使用nvidia-smi命令可以查看

PS C:\Users\gouff> nvidia-smi
Sat Jul 15 16:24:16 2023
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 536.40                 Driver Version: 536.40       CUDA Version: 12.2     |

可以看出驱动的版本是536.40，cuda的版本是12.2

step2. 配置wsl (略过)，保障是wsl2

step3.在wsl的ubuntu22里安装gcc

sudo apt-get install -y gcc --fix-missing

step4.安装wsl版本的cuda

CUDA Toolkit 12.2 Downloads | NVIDIA Developer

按照下边的命令安装

step5.环境变量设置

NVIDIA CUDA Installation Guide for Linux

查看

step6.验证

输入下面两个命令，如果成功，那就ok了。

nvcc -V
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2023 NVIDIA Corporation
Built on Tue_Jun_13_19:16:58_PDT_2023
Cuda compilation tools, release 12.2, V12.2.91
Build cuda_12.2.r12.2/compiler.32965470_0

nvidia-smi
Sat Jul 15 18:23:09 2023
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.54.06              Driver Version: 536.40       CUDA Version: 12.2     |
|-----------------------------------------+----------------------+----------------------+
| 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  NVIDIA GeForce RTX 4070        On  | 00000000:01:00.0  On |                  N/A |
|  0%   30C    P8               6W / 215W |   3245MiB / 12282MiB |      1%      Default |
|                                         |                      |                  N/A |
+-----------------------------------------+----------------------+----------------------+

step7.说明

如果执行cuda程序出错，请补充下边环境变量

export LD_LIBRARY_PATH=/usr/lib/wsl/lib:$LD_LIBRARY_PATH