<aside> 🧸 CUDA相关内容

</aside>

【Install and Update】
```
nvcc -V
```
```
nvidia-smi
```
```
C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA
```
```
C:\\Program Files\\NVIDIA Corporation
```
<aside>

CUDA_PATH 这个系统级别的环境变量，需要手动指向所需的CUDA，以供 “构建系统”参考；

PATH 中需要加入实际CUDA路径，以运行nvcc -V ，注意这里的顺序也影响实际解析结果

</aside>

NVIDIA CUDA Toolkit 12.1 Downloads

https://developer.nvidia.com/cuda-downloads?target_os=Windows&target_arch=x86_64&target_version=11&target_type=exe_local
- 选择相应<distro> <version> <architecture>安装：https://developer.nvidia.com/cuda-downloads?target_os=Linux&target_arch=x86_64&Distribution=Debian&target_version=12&target_type=deb_local
```
software-properties-common
```
- 更新CUDA Software及相关驱动、软件：
  - https://docs.nvidia.com/cuda/cuda-installation-guide-microsoft-windows/#download-the-nvidia-cuda-toolkit （在Graphical Installation可以选择安装全部CUDA Software）
  - 图像驱动程序、physX 系统软件和 HD音频也可以在GeForce/ APP中更新驱动。
  - NVIDIA app
<aside> 🚧

Once a Windows NVIDIA GPU driver is installed on the system, CUDA becomes available within WSL 2. users must not install any NVIDIA GPU Linux driver within WSL 2.

https://developer.nvidia.com/cuda-downloads?target_os=Linux&target_arch=x86_64&Distribution=WSL-Ubuntu&target_version=2.0&target_type=deb_local

</aside>

<aside>

cuDSS 也是 NVIDIA 系列产品 cuDSS

</aside>

PyTorch

uv pip  install torch torchvision torchaudio --index-url <https://download.pytorch.org/whl/>
cu126

import torch
torch.cuda.is_available()