Skip to content

Latest commit

 

History

History
92 lines (67 loc) · 2.57 KB

README_mac_m1.md

File metadata and controls

92 lines (67 loc) · 2.57 KB

Update 2023/07/22

openvino now has official support for mac M1, so just install it with brew.

brew install openvino

How to build openvino with apple M1

openvino arm plugin build need SCons

brew install scons

clone source and build

cd ~

git clone https://github.com/openvinotoolkit/openvino.git
git clone https://github.com/openvinotoolkit/openvino_contrib.git

cd openvino         && git submodule update --init --recursive && cd ..
cd openvino_contrib && git submodule update --init --recursive && cd ..

mkdir openvino/build && cd openvino/build
# IE_EXTRA_MODULES must be absolute path.
cmake -DCMAKE_BUILD_TYPE=Release -DENABLE_PYTHON=OFF -DIE_EXTRA_MODULES=~/openvino_contrib/modules/arm_plugin ..
make -j8
sudo make install

check supported devices.

➜  build git:(master) ../bin/arm64/Release/hello_query_device
[ INFO ] OpenVINO Runtime version ......... 2022.3.0
[ INFO ] Build ........... 2022.3.0-000-d9d4b6d89ba
[ INFO ]
[ INFO ] Available devices:
[ INFO ] CPU
[ INFO ] 	SUPPORTED_PROPERTIES:
[ INFO ] 		Immutable: SUPPORTED_METRICS : SUPPORTED_METRICS SUPPORTED_CONFIG_KEYS RANGE_FOR_ASYNC_INFER_REQUESTS RANGE_FOR_STREAMS
[ INFO ] 		Immutable: SUPPORTED_CONFIG_KEYS : LP_TRANSFORMS_MODE DUMP_GRAPH PERF_COUNT CPU_THROUGHPUT_STREAMS CPU_BIND_THREAD CPU_THREADS_NUM CPU_THREADS_PER_STREAM NUM_STREAMS INFERENCE_NUM_THREADS AFFINITY
[ INFO ] 		Mutable: PERF_COUNT : YES
[ INFO ] 		Immutable: AVAILABLE_DEVICES : NEON
[ INFO ] 		Immutable: FULL_DEVICE_NAME : arm_compute::NEON
[ INFO ] 		Immutable: OPTIMIZATION_CAPABILITIES : FP16 FP32
[ INFO ] 		Immutable: RANGE_FOR_ASYNC_INFER_REQUESTS : 1 10 1
[ INFO ] 		Immutable: RANGE_FOR_STREAMS : 1 10
[ INFO ] 		Mutable: CPU_THROUGHPUT_STREAMS : 1
[ INFO ] 		Mutable: CPU_BIND_THREAD : NO
[ INFO ] 		Mutable: CPU_THREADS_NUM : 0
[ INFO ] 		Mutable: CPU_THREADS_PER_STREAM : 10
[ INFO ] 		Mutable: NUM_STREAMS : 1
[ INFO ] 		Mutable: INFERENCE_NUM_THREADS : 0
[ INFO ] 		Mutable: AFFINITY : NONE
[ INFO ]

It will install dynamic libraries to /usr/local/runtime/lib/arm64/Release

link it to system lib dir

sudo ln -s /usr/local/runtime/lib/arm64/Release/libopenvino.2230.dylib /usr/local/lib/

now we can add run path to the our server binary.

cd face-recognition-service

# run server
./build/server/bin/FaceRecognitionServer --config_path ./contrib/server.json

# run test
./build/test/bin/FaceRecognitionTests

Conan

export CC=/opt/homebrew/opt/llvm/bin/clang ; export CXX=/opt/homebrew/opt/llvm/bin/clang++

conan install --build=missing --profile conan/conanprofile.m1  -if build ./conan