Hello ORT(OnnxRumTime)
An ONNX(Open Neural Network Exchange) Runtime Demo
- ORT DOC: https://onnxruntime.ai/docs
- ORT Github: https://github.com/microsoft/onnxruntime
- ORT Maven Repo: https://mvnrepository.com/artifact/com.microsoft.onnxruntime
- ORT Architecture:
- ORT Workflow:
1 CUDA
- Tesla 特斯拉
- Fermi 费米
- Kepler 开普勒
- Maxwell 麦克斯韦
- Pascal 帕斯卡
- Volta 伏打
- Turing 图灵
- Ampere 安培
- Ada Lovelace 阿达·洛芙莱斯
- Hopper 格蕾丝·赫柏
- Blackwell 戴维·布莱克维尔
- https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html#requirements
- https://docs.nvidia.com/deeplearning/cudnn/latest/installation/overview.html
ONNX Runtime | CUDA(Compute Unified Device Architecture) | cuDNN(CUDA Deep Neural Network library) | |
---|---|---|---|
url | https://github.com/microsoft/onnxruntime/tags | https://developer.nvidia.com/cuda-toolkit-archive | https://developer.nvidia.com/rdp/cudnn-archive |
version | 1.18 | CUDA Toolkit 11.8.0 (October 2022) | cuDNN v8.9.2 (June 1st, 2023), for CUDA 11.x |
version | 1.18 | CUDA Toolkit 12.5.0 (May 2024) | cuDNN v8.9.7 (December 5th, 2023), for CUDA 12.x |
CUDA_PATH=D:\garden\cuda12.5
CUDNN_PATH=D:\garden\cudnn8.9.7.29_cuda12
PATH=%CUDA_PATH%\bin;%CUDA_PATH%\libnvvp;%CUDNN_PATH%\bin;...
cp "C:\Program Files\NVIDIA Corporation\Nsight Systems 2024.2.3\host-windows-x64\zlib.dll" D:\garden\cuda12.5\bin
env | grep CUDA
CUDA Driver API 版本
nvidia-smi
CUDA Runtime API 版本
nvcc -V
每隔5秒刷新一次GPU使用情况
watch -n 5 nvidia-smi
https://developer.apple.com/documentation/coreml
Which devices have an ANE?
https://github.com/hollance/neural-engine/blob/master/docs/supported-devices.md
UnsatisfiedLink no onnxruntime in java.library.path
reload the dependency jar file and check it:
no cudart in java.library.path
-Djava.library.path=D:\\coding\\cuda\\11.8\\lib\\x64