How to Implement Custom CPU Operations

The primary means of the performance of the CPU codepath in the Inference Engine is the Intel® Math Kernel Library for Deep Neural Networks (Intel® MKL-DNN), and new CPU kernels extend the Inference Engine plugin for the Intel MKL-DNN. Implementing the InferenceEngine::ILayerExecImpl defines a general CPU-side extension. There are no Intel MKL-DNN specifics in the way you need to implement a kernel.

Implementation Class

All custom kernels for the CPU plugin should be inherited from the InferenceEngine::ILayerExecImpl interface. Based on that, declaration of a kernel implementation class can look as follows:

class OpImplementation : public InferenceEngine::ILayerExecImpl {
explicit OpImplementation(const std::shared_ptr<ngraph::Node>& node);
InferenceEngine::StatusCode getSupportedConfigurations(std::vector<InferenceEngine::LayerConfig> &conf,
InferenceEngine::ResponseDesc *resp) noexcept override;
InferenceEngine::ResponseDesc *resp) noexcept override;
InferenceEngine::StatusCode execute(std::vector<InferenceEngine::Blob::Ptr> &inputs,
std::vector<InferenceEngine::Blob::Ptr> &outputs,
InferenceEngine::ResponseDesc *resp) noexcept override;
int64_t add;
ngraph::Shape inShape;
ngraph::Shape outShape;
std::string error;
This class provides interface for the implementation with the custom execution code.
Definition: ie_iextension.h:104
virtual StatusCode getSupportedConfigurations(std::vector< LayerConfig > &conf, ResponseDesc *resp) noexcept=0
Gets all supported configurations for the current layer.
virtual StatusCode execute(std::vector< Blob::Ptr > &inputs, std::vector< Blob::Ptr > &outputs, ResponseDesc *resp) noexcept=0
Execute method.
virtual StatusCode init(LayerConfig &config, ResponseDesc *resp) noexcept=0
Initializes the implementation.
This enum contains codes for all possible return values of the interface functions.
Definition: ie_common.h:224
Node add(NodeInput left_node, NodeInput right_node, str auto_broadcast="NUMPY", Optional[str] name=None)
This structure describes Layer configuration.
Definition: ie_iextension.h:68
Represents detailed information for an error.
Definition: ie_common.h:248

Class Fields

The provided implementation has several fields:

  • add of the type int64_t is an attribute of a custom operation.
  • inShape of the type ngraph::Shape is an input shape.
  • outShape of the type ngraph::Shape is an output shape.
  • error of the type std::string is a field to handle errors from a constructor.

Constructor of Implementation

An implementation constructor checks parameters of an nGraph operation, stores required attributes, and stores an error message in the case of an error.

OpImplementation::OpImplementation(const std::shared_ptr<ngraph::Node> &node) {
try {
auto castedNode = std::dynamic_pointer_cast<Operation>(node);
if (!castedNode)
THROW_IE_EXCEPTION << "Cannot create implementation for unknown operation!";
if (castedNode->inputs().size() != 1 || castedNode->outputs().size() != 1)
THROW_IE_EXCEPTION << "Cannot create implementation for operation with incorrect number of inputs or outputs!";
if (castedNode->get_input_partial_shape(0).is_dynamic() || castedNode->get_output_partial_shape(0).is_dynamic())
THROW_IE_EXCEPTION << "Cannot create implementation for op with dynamic shapes!";
if (castedNode->get_input_shape(0).size() != 4 || castedNode->get_output_shape(0).size() != 4)
THROW_IE_EXCEPTION << "Operation supports only 4d tensors for input and output.";
if (castedNode->get_input_element_type(0) != ngraph::element::f32 || castedNode->get_output_element_type(0) != ngraph::element::f32)
THROW_IE_EXCEPTION << "Operation supports only FP32 tensors.";
add = castedNode->getAddAttr();
inShape = castedNode->get_input_shape(0);
outShape = castedNode->get_output_shape(0);
} catch (InferenceEngine::details::InferenceEngineException& ex) {
error = ex.what();
A macro used to throw general exception with a description.
Definition: ie_exception.hpp:25


InferenceEngine::ILayerExecImpl::getSupportedConfigurations method returns all supported configuration formats (input/output tensor layouts) for your implementation. To specify formats of data, use InferenceEngine::TensorDesc. Refer to the Memory Primitives section for instructions.

InferenceEngine::StatusCode OpImplementation::getSupportedConfigurations(std::vector<InferenceEngine::LayerConfig> &conf,
auto createConfig = [](const InferenceEngine::SizeVector inShape, const InferenceEngine::SizeVector& outShape, bool planar) {
config.dynBatchSupport = false;
InferenceEngine::SizeVector order = {0, 1, 2, 3};
// Allow any offset before data
size_t offset((std::numeric_limits<size_t>::max)());
if (planar) {
inData.desc = InferenceEngine::TensorDesc(InferenceEngine::Precision::FP32, inShape, {inShape, order, offset});
outData.desc = InferenceEngine::TensorDesc(InferenceEngine::Precision::FP32, outShape, {outShape, order, offset});
} else {
// Add blocked (nChw8c) format
auto div_up = [](const int a, const int b) -> int {
if (!b)
return 0;
return (a + b - 1) / b;
InferenceEngine::SizeVector inBlkDims = inShape;
inBlkDims[1] = div_up(inBlkDims[1], 8);
InferenceEngine::SizeVector outBlkDims = outShape;
outBlkDims[1] = div_up(outBlkDims[1], 8);
inData.desc = InferenceEngine::TensorDesc(InferenceEngine::Precision::FP32, inShape, {inBlkDims, order, offset});
outData.desc = InferenceEngine::TensorDesc(InferenceEngine::Precision::FP32, outShape, {outBlkDims, order, offset});
return config;
if (!error.empty()) {
if (resp) {
strncpy(resp->msg, error.c_str(), sizeof(resp->msg) - 1);
resp->msg[sizeof(resp->msg)-1] = 0;
return InferenceEngine::GENERAL_ERROR;
// Add planar format
conf.emplace_back(createConfig(inShape, outShape, true));
// Add blocked format nChw8c
conf.emplace_back(createConfig(inShape, outShape, false));
return InferenceEngine::OK;
@ FP32
Definition: ie_precision.hpp:29
This class defines Tensor description.
Definition: ie_layouts.h:158
std::vector< size_t > SizeVector
Represents tensor size.
Definition: ie_common.h:27
This structure describes data configuration.
Definition: ie_iextension.h:48
TensorDesc desc
Format of memory descriptor.
Definition: ie_iextension.h:52


InferenceEngine::ILayerExecImpl::init method gets a runtime-selected configuration from a vector that is populated from the getSupportedConfigurations method and checks the parameters:

try {
if (config.inConfs.size() != 1 || config.outConfs.size() != 1) {
THROW_IE_EXCEPTION << "Operation cannot be initialized with incorrect number of inputs/outputs!";
if (config.inConfs[0].desc.getDims().size() != 4 || config.outConfs[0].desc.getDims().size() != 4) {
THROW_IE_EXCEPTION << "Operation can be initialized only with 4d input/output tensors!";
if (config.outConfs[0].desc.getPrecision() != InferenceEngine::Precision::FP32 ||
config.inConfs[0].desc.getPrecision() != InferenceEngine::Precision::FP32) {
THROW_IE_EXCEPTION << "Operation supports only FP32 precisions!";
} catch (InferenceEngine::details::InferenceEngineException& ex) {
if (resp) {
strncpy(resp->msg, error.c_str(), sizeof(resp->msg) - 1);
resp->msg[sizeof(resp->msg)-1] = 0;
return InferenceEngine::GENERAL_ERROR;
return InferenceEngine::OK;


InferenceEngine::ILayerExecImpl::execute method accepts and processes the actual tenors as input/output blobs:

InferenceEngine::StatusCode OpImplementation::execute(std::vector<InferenceEngine::Blob::Ptr> &inputs,
std::vector<InferenceEngine::Blob::Ptr> &outputs,
const float* src_data = inputs[0]->cbuffer().as<const float *>() + inputs[0]->getTensorDesc().getBlockingDesc().getOffsetPadding();
float *dst_data = outputs[0]->buffer().as<float *>() + outputs[0]->getTensorDesc().getBlockingDesc().getOffsetPadding();
for (size_t i = 0; i < inputs[0]->size(); i++) {
dst_data[i] = src_data[i] + add;
return InferenceEngine::OK;

Register Implementation in Extension Class

To register custom kernel implementation in the Extension class, implement the following methods:


InferenceEngine::IExtension::getImplTypes returns a vector of implementation types for an operation.

std::vector<std::string> Extension::getImplTypes(const std::shared_ptr<ngraph::Node> &node) {
if (std::dynamic_pointer_cast<Operation>(node)) {
return {"CPU"};
if (std::dynamic_pointer_cast<FFTOp>(node)) {
return {"CPU"};
return {};


InferenceEngine::IExtension::getImplementation returns the kernel implementation with a specified type for an operation.

InferenceEngine::ILayerImpl::Ptr Extension::getImplementation(const std::shared_ptr<ngraph::Node> &node, const std::string &implType) {
if (implType == "CPU") {
if (std::dynamic_pointer_cast<Operation>(node)) {
return std::make_shared<OpImplementation>(node);
if (std::dynamic_pointer_cast<FFTOp>(node) && implType == "CPU") {
return std::make_shared<FFTImpl>(node);
return nullptr;
std::shared_ptr< ILayerImpl > Ptr
A shared pointer to the ILayerImpl interface.
Definition: ie_iextension.h:92

Load Extension with Executable Kernels to Plugin

Use the AddExtension method of the general plugin interface to load your primitives:

// Load CPU extension as a shared library
auto extension_ptr = make_so_pointer<InferenceEngine::IExtension>("<shared lib path>");
// Add extension to the CPU device
core.AddExtension(extension_ptr, "CPU");
This class represents Inference Engine Core entity.
Definition: ie_core.hpp:29