OpenVINO™ Execution Provider

Accelerate ONNX models on Intel CPUs, GPUs with Intel OpenVINO™ Execution Provider. Please refer to this page for details on the Intel hardware supported.

Contents

Install

Pre-built packages and Docker images are published for OpenVINO™ Execution Provider for ONNX Runtime by Intel for each release.

Requirements

ONNX Runtime OpenVINO™ Execution Provider is compatible with three lastest releases of OpenVINO™.

ONNX Runtime OpenVINO™ Notes
1.17.1 2023.3 Details
1.17.1 2023.2 Details
1.16.0 2023.1 Details
1.15.0 2023.0 Details
1.14.0 2022.3 Details

Build

For build instructions, please see the BUILD page.

Usage

Set OpenVINO™ Environment for Python

Please download onnxruntime-openvino python packages from PyPi.org:

pip install onnxruntime-openvino
  • Windows

    To enable OpenVINO™ Execution Provider with ONNX Runtime on Windows it is must to set up the OpenVINO™ Environment Variables using the full installer package of OpenVINO™. Initialize the OpenVINO™ environment by running the setupvars script as shown below. This is a required step:

        C:\ <openvino_install_directory>\setupvars.bat
    
  • Linux

    OpenVINO™ Execution Provider with Onnx Runtime on Linux, installed from PyPi.org comes with prebuilt OpenVINO™ libs and supports flag CXX11_ABI=0. So there is no need to install OpenVINO™ separately.

    But if there is need to enable CX11_ABI=1 flag of OpenVINO, build Onnx Runtime python wheel packages from source. For build instructions, please see the BUILD page. OpenVINO™ Execution Provider wheels on Linux built from source will not have prebuilt OpenVINO™ libs so we must set the OpenVINO™ Environment Variable using the full installer package of OpenVINO™:

    ```
    $ source <openvino_install_directory>/setupvars.sh
    ```
    

Set OpenVINO™ Environment for C++

For Running C++/C# ORT Samples with the OpenVINO™ Execution Provider it is must to set up the OpenVINO™ Environment Variables using the full installer package of OpenVINO™. Initialize the OpenVINO™ environment by running the setupvars script as shown below. This is a required step:

  • For Windows run:
     C:\ <openvino_install_directory>\setupvars.bat
    
  • For Linux run:
     $ source <openvino_install_directory>/setupvars.sh
    

    Note: If you are using a dockerfile to use OpenVINO™ Execution Provider, sourcing OpenVINO™ won’t be possible within the dockerfile. You would have to explicitly set the LD_LIBRARY_PATH to point to OpenVINO™ libraries location. Refer our dockerfile.

Set OpenVINO™ Environment for C#

To use csharp api for openvino execution provider create a custom nuget package. Follow the instructions here to install prerequisites for nuget creation. Once prerequisites are installed follow the instructions to build openvino execution provider and add an extra flag --build_nuget to create nuget packages. Two nuget packages will be created Microsoft.ML.OnnxRuntime.Managed and Microsoft.ML.OnnxRuntime.Openvino.

Features

OpenCL queue throttling for GPU devices

Enables OpenCL queue throttling for GPU devices. Reduces CPU utilization when using GPUs with OpenVINO EP.

Model caching

OpenVINO™ supports model caching.

From OpenVINO™ 2023.1 version, model caching feature is supported on CPU, GPU along with kernel caching on iGPU, dGPU.

This feature enables users to save and load the blob file directly on to the hardware device target and perform inference with improved Inference Latency.

Kernel Caching on iGPU and dGPU:

This feature also allows user to save kernel caching as cl_cache files for models with dynamic input shapes. These cl_cache files can be loaded directly onto the iGPU/dGPU hardware device target and inferencing can be performed.

Enabling Model Caching via Runtime options using c++/python API’s.

This flow can be enabled by setting the runtime config option ‘cache_dir’ specifying the path to dump and load the blobs (CPU, iGPU, dGPU) or cl_cache(iGPU, dGPU) while using the c++/python API’S.

Refer to Configuration Options for more information about using these runtime options.

Support for INT8 Quantized models

Int8 models are supported on CPU and GPU.

Support for Weights saved in external files

OpenVINO™ Execution Provider now supports ONNX models that store weights in external files. It is especially useful for models larger than 2GB because of protobuf limitations.

See the OpenVINO™ ONNX Support documentation.

Converting and Saving an ONNX Model to External Data: Use the ONNX API’s.documentation.

Example:

import onnx
onnx_model = onnx.load("model.onnx") # Your model in memory as ModelProto
onnx.save_model(onnx_model, 'saved_model.onnx', save_as_external_data=True, all_tensors_to_one_file=True, location='data/weights_data', size_threshold=1024, convert_attribute=False)

Note:

  1. In the above script, model.onnx is loaded and then gets saved into a file called ‘saved_model.onnx’ which won’t have the weights but this new onnx model now will have the relative path to where the weights file is located. The weights file ‘weights_data’ will now contain the weights of the model and the weights from the original model gets saved at /data/weights_data.

  2. Now, you can use this ‘saved_model.onnx’ file to infer using your sample. But remember, the weights file location can’t be changed. The weights have to be present at /data/weights_data

  3. Install the latest ONNX Python package using pip to run these ONNX Python API’s successfully.

Support for IO Buffer Optimization

To enable IO Buffer Optimization we have to set OPENCL_LIBS, OPENCL_INCS environment variables before build. For IO Buffer Optimization, the model must be fully supported on OpenVINO™ and we must provide in the remote context cl_context void pointer as C++ Configuration Option. We can provide cl::Buffer address as Input using GPU Memory Allocator for input and output.

Example:

//Set up a remote context
cl::Context _context;
.....
// Set the context through openvino options
std::unordered_map<std::string, std::string> ov_options;
ov_options[context] = std::to_string((unsigned long long)(void *) _context.get());
.....
//Define the Memory area
Ort::MemoryInfo info_gpu("OpenVINO_GPU", OrtAllocatorType::OrtDeviceAllocator, 0, OrtMemTypeDefault);
//Create a shared buffer , fill in with data
cl::Buffer shared_buffer(_context, CL_MEM_READ_WRITE, imgSize, NULL, &err);
....
//Cast it to void*, and wrap it as device pointer for Ort::Value
void *shared_buffer_void = static_cast<void *>(&shared_buffer);
Ort::Value inputTensors = Ort::Value::CreateTensor(
        info_gpu, shared_buffer_void, imgSize, inputDims.data(),
        inputDims.size(), ONNX_TENSOR_ELEMENT_DATA_TYPE_FLOAT);

Multi-threading for OpenVINO™ Execution Provider

OpenVINO™ Execution Provider for ONNX Runtime enables thread-safe deep learning inference

Multi streams for OpenVINO™ Execution Provider

OpenVINO™ Execution Provider for ONNX Runtime allows multiple stream execution for difference performance requirements part of API 2.0

Auto-Device Execution for OpenVINO EP

Use AUTO:<device 1><device 2>.. as the device name to delegate selection of an actual accelerator to OpenVINO™. Auto-device internally recognizes and selects devices from CPU, integrated GPU and discrete Intel GPUs (when available) depending on the device capabilities and the characteristic of CNN models, for example, precisions. Then Auto-device assigns inference requests to the selected device.

From the application point of view, this is just another device that handles all accelerators in full system.

For more information on Auto-Device plugin of OpenVINO™, please refer to the Intel OpenVINO™ Auto Device Plugin.

Heterogeneous Execution for OpenVINO™ Execution Provider

The heterogeneous execution enables computing for inference on one network on several devices. Purposes to execute networks in heterogeneous mode:

  • To utilize accelerator’s power and calculate the heaviest parts of the network on the accelerator and execute unsupported layers on fallback devices like the CPU to utilize all available hardware more efficiently during one inference.

For more information on Heterogeneous plugin of OpenVINO™, please refer to the Intel OpenVINO™ Heterogeneous Plugin.

Multi-Device Execution for OpenVINO EP

Multi-Device plugin automatically assigns inference requests to available computational devices to execute the requests in parallel. Potential gains are as follows:

  • Improved throughput that multiple devices can deliver (compared to single-device execution)
  • More consistent performance, since the devices can now share the inference burden (so that if one device is becoming too busy, another device can take more of the load)

For more information on Multi-Device plugin of OpenVINO™, please refer to the Intel OpenVINO™ Multi Device Plugin.

Configuration Options

OpenVINO™ Execution Provider can be configured with certain options at runtime that control the behavior of the EP. These options can be set as key-value pairs as below:-

Python API

Key-Value pairs for config options can be set using InferenceSession API as follow:-

session = onnxruntime.InferenceSession(<path_to_model_file>, providers=['OpenVINOExecutionProvider'], provider_options=[{Key1 : Value1, Key2 : Value2, ...}])

Note that the releases from (ORT 1.10) will require explicitly setting the providers parameter if you want to use execution providers other than the default CPU provider (as opposed to the current behavior of providers getting set/registered by default based on the build flags) when instantiating InferenceSession.

C/C++ API 2.0

The session configuration options are passed to SessionOptionsAppendExecutionProvider API as shown in an example below for GPU device type:

std::unordered_map<std::string, std::string> options;
options[device_type] = "GPU_FP32";
options[device_id] = "";
options[num_of_threads] = "8";
options[num_streams] = "8";
options[cache_dir] = "";
options[context] = "0x123456ff";
options[enable_opencl_throttling] = "false";
session_options.AppendExecutionProvider("OpenVINO", options);

C/C++ Legacy API

The session configuration options are passed to SessionOptionsAppendExecutionProvider_OpenVINO() API as shown in an example below for GPU device type:

OrtOpenVINOProviderOptions options;
options.device_type = "GPU_FP32";
options.device_id = "";
options.num_of_threads = 8;
options.cache_dir = "";
options.context = 0x123456ff;
options.enable_opencl_throttling = false;
SessionOptions.AppendExecutionProvider_OpenVINO(session_options, &options);

Onnxruntime Graph level Optimization

OpenVINO™ backend performs hardware, dependent as well as independent optimizations on the graph to infer it on the target hardware with best possible performance. In most cases it has been observed that passing the ONNX input graph as is without explicit optimizations would lead to best possible optimizations at kernel level by OpenVINO™. For this reason, it is advised to turn off high level optimizations performed by ONNX Runtime for OpenVINO™ Execution Provider. This can be done using SessionOptions() as shown below:-

  • Python API

     options = onnxruntime.SessionOptions()
     options.graph_optimization_level = onnxruntime.GraphOptimizationLevel.ORT_DISABLE_ALL
     sess = onnxruntime.InferenceSession(<path_to_model_file>, options)
    
  • C/C++ API

     SessionOptions::SetGraphOptimizationLevel(ORT_DISABLE_ALL);
    

Summary of options

The following table lists all the available configuration options for API 2.0 and the Key-Value pairs to set them:

Key Key type Allowable Values Value type Description  
device_type string CPU_FP32, CPU_FP16, GPU_FP32, GPU_FP16, GPU.0_FP32, GPU.1_FP32, GPU.0_FP16, GPU.1_FP16 based on the avaialable GPUs, Any valid Hetero combination, Any valid Multi or Auto devices combination string Overrides the accelerator hardware type and precision with these values at runtime. If this option is not explicitly set, default hardware and precision specified during build time is used. Overrides the accelerator hardware type and precision with these values at runtime. If this option is not explicitly set, default hardware and precision specified during build time is used.
device_id string Any valid OpenVINO device ID string Selects a particular hardware device for inference. The list of valid OpenVINO device ID’s available on a platform can be obtained either by Python API (onnxruntime.capi._pybind_state.get_available_openvino_device_ids()) or by OpenVINO C/C++ API. If this option is not explicitly set, an arbitrary free device will be automatically selected by OpenVINO runtime.  
num_of_threads string Any unsigned positive number other than 0 size_t Overrides the accelerator default value of number of threads with this value at runtime. If this option is not explicitly set, default value of 8 during build time will be used for inference.  
num_streams string Any unsigned positive number other than 0 size_t Overrides the accelerator default streams with this value at runtime. If this option is not explicitly set, default value of 1, performance for latency is used during build time will be used for inference.  
cache_dir string Any valid string path on the hardware target string Explicitly specify the path to save and load the blobs enabling model caching feature.  
context string OpenCL Context void* This option is only available when OpenVINO EP is built with OpenCL flags enabled. It takes in the remote context i.e the cl_context address as a void pointer.  
enable_opencl_throttling string True/False boolean This option enables OpenCL queue throttling for GPU devices (reduces CPU utilization when using GPU).  

Valid Hetero or Multi or Auto Device combinations: HETERO:,,... The can be any of these devices from this list ['CPU','GPU']

A minimum of two DEVICE_TYPE’S should be specified for a valid HETERO or Multi-Device Build.

Example: HETERO:GPU,CPU AUTO:GPU,CPU MULTI:GPU,CPU

Support Coverage

ONNX Layers supported using OpenVINO

The table below shows the ONNX layers supported and validated using OpenVINO™ Execution Provider.The below table also lists the Intel hardware support for each of the layers. CPU refers to Intel® Atom, Core, and Xeon processors. GPU refers to the Intel Integrated Graphics. Intel Discrete Graphics

ONNX Layers CPU GPU  
Abs Yes Yes  
Acos Yes Yes  
Acosh Yes Yes  
Add Yes Yes  
And Yes Yes  
ArgMax Yes Yes  
ArgMin Yes Yes  
Asin Yes Yes  
Asinh Yes Yes  
Atan Yes Yes  
Atanh Yes Yes  
AveragePool Yes Yes  
BatchNormalization Yes Yes  
BitShift Yes No  
Ceil Yes Yes  
Celu Yes Yes  
Cast Yes Yes  
Clip Yes Yes  
Concat Yes Yes  
Constant Yes Yes  
ConstantOfShape Yes Yes  
Conv Yes Yes  
ConvInteger Yes Yes  
ConvTranspose Yes Yes  
Cos Yes Yes  
Cosh Yes Yes  
CumSum Yes Yes  
DepthToSpace Yes Yes  
DequantizeLinear Yes Yes  
Div Yes Yes  
Dropout Yes Yes  
Einsum Yes Yes  
Elu Yes Yes  
Equal Yes Yes  
Erf Yes Yes  
Exp Yes Yes  
Expand Yes Yes  
EyeLike Yes No  
Flatten Yes Yes  
Floor Yes Yes  
Gather Yes Yes  
GatherElements No No  
GatherND Yes Yes  
Gemm Yes Yes  
GlobalAveragePool Yes Yes  
GlobalLpPool Yes Yes  
GlobalMaxPool Yes Yes  
Greater Yes Yes Yes
GreaterOrEqual Yes Yes  
GridSample Yes No  
HardMax Yes Yes  
HardSigmoid Yes Yes  
Identity Yes Yes  
If Yes Yes  
ImageScaler Yes Yes  
InstanceNormalization Yes Yes  
LeakyRelu Yes Yes  
Less Yes Yes  
LessOrEqual Yes Yes  
Log Yes Yes  
LogSoftMax Yes Yes  
Loop Yes Yes  
LRN Yes Yes  
LSTM Yes Yes  
MatMul Yes Yes  
MatMulInteger Yes No  
Max Yes Yes  
MaxPool Yes Yes  
Mean Yes Yes  
MeanVarianceNormalization Yes Yes  
Min Yes Yes  
Mod Yes Yes  
Mul Yes Yes  
Neg Yes Yes  
NonMaxSuppression Yes Yes  
NonZero Yes No  
Not Yes Yes  
OneHot Yes Yes  
Or Yes Yes  
Pad Yes Yes  
Pow Yes Yes  
PRelu Yes Yes  
QuantizeLinear Yes Yes  
QLinearMatMul Yes No  
Range Yes Yes  
Reciprocal Yes Yes  
ReduceL1 Yes Yes  
ReduceL2 Yes Yes  
ReduceLogSum Yes Yes  
ReduceLogSumExp Yes Yes  
ReduceMax Yes Yes  
ReduceMean Yes Yes  
ReduceMin Yes Yes  
ReduceProd Yes Yes  
ReduceSum Yes Yes  
ReduceSumSquare Yes Yes  
Relu Yes Yes  
Reshape Yes Yes  
Resize Yes Yes  
ReverseSequence Yes Yes  
RoiAlign Yes Yes  
Round Yes Yes  
Scatter Yes Yes  
ScatterElements Yes Yes  
ScatterND Yes Yes  
Selu Yes Yes  
Shape Yes Yes  
Shrink Yes Yes  
Sigmoid Yes Yes  
Sign Yes Yes  
Sin Yes Yes  
Sinh Yes No  
SinFloat No No  
Size Yes Yes  
Slice Yes Yes  
Softmax Yes Yes  
Softplus Yes Yes  
Softsign Yes Yes  
SpaceToDepth Yes Yes  
Split Yes Yes  
Sqrt Yes Yes  
Squeeze Yes Yes  
Sub Yes Yes  
Sum Yes Yes  
Softsign Yes No  
Tan Yes Yes  
Tanh Yes Yes  
ThresholdedRelu Yes Yes  
Tile Yes Yes  
TopK Yes Yes  
Transpose Yes Yes  
Unsqueeze Yes Yes  
Upsample Yes Yes  
Where Yes Yes  
Xor Yes Yes  

Topology Support

Below topologies from ONNX open model zoo are fully supported on OpenVINO™ Execution Provider and many more are supported through sub-graph partitioning

Image Classification Networks

MODEL NAME CPU GPU
bvlc_alexnet Yes Yes
bvlc_googlenet Yes Yes
bvlc_reference_caffenet Yes Yes
bvlc_reference_rcnn_ilsvrc13 Yes Yes
emotion ferplus Yes Yes
densenet121 Yes Yes
inception_v1 Yes Yes
inception_v2 Yes Yes
mobilenetv2 Yes Yes
resnet18v2 Yes Yes
resnet34v2 Yes Yes
resnet101v2 Yes Yes
resnet152v2 Yes Yes
resnet50 Yes Yes
resnet50v2 Yes Yes
shufflenet Yes Yes
squeezenet1.1 Yes Yes
vgg19 Yes Yes
zfnet512 Yes Yes
mxnet_arcface Yes Yes

Image Recognition Networks

MODEL NAME CPU GPU
mnist Yes Yes

Object Detection Networks

MODEL NAME CPU GPU
tiny_yolov2 Yes Yes
yolov3 Yes Yes
tiny_yolov3 Yes Yes
mask_rcnn Yes No
faster_rcnn Yes No
yolov4 Yes Yes
yolov5 Yes Yes
yolov7 Yes Yes
tiny_yolov7 Yes Yes

Image Manipulation Networks

MODEL NAME CPU GPU
mosaic Yes Yes
candy Yes Yes
cgan Yes Yes
rain_princess Yes Yes
pointilism Yes Yes
udnie Yes Yes

Natural Language Processing Networks

MODEL NAME CPU GPU
bert-squad Yes Yes
bert-base-cased Yes Yes
bert-base-chinese Yes Yes
bert-base-japanese-char Yes Yes
bert-base-multilingual-cased Yes Yes
bert-base-uncased Yes Yes
distilbert-base-cased Yes Yes
distilbert-base-multilingual-cased Yes Yes
distilbert-base-uncased Yes Yes
distilbert-base-uncased-finetuned-sst-2-english Yes Yes
gpt2 Yes Yes
roberta-base Yes Yes
roberta-base-squad2 Yes Yes
t5-base Yes Yes
twitter-roberta-base-sentiment Yes Yes
xlm-roberta-base Yes Yes

Note: We have added support for INT8 models, quantized with Neural Network Compression Framework (NNCF). To know more about NNCF refer here.

OpenVINO™ Execution Provider Samples Tutorials

In order to showcase what you can do with the OpenVINO™ Execution Provider for ONNX Runtime, we have created a few samples that shows how you can get that performance boost you’re looking for with just one additional line of code.

Python API

Object detection with tinyYOLOv2 in Python

Object detection with YOLOv4 in Python

C/C++ API

Image classification with Squeezenet in CPP

Csharp API

Object detection with YOLOv3 in C#

Blogs/Tutorials

Overview of OpenVINO Execution Provider for ONNX Runtime

OpenVINO Execution Provider

Tutorial on how to use OpenVINO™ Execution Provider for ONNX Runtime Docker Containers

Docker Containers

Tutorial on how to use OpenVINO™ Execution Provider for ONNX Runtime python wheel packages

Python Pip Wheel Packages