C++ API Reference for Intel® Data Analytics Acceleration Library 2019

List of all members
AlgorithmDispatchLayerContainer< mode, DAAL_KERNEL_AVX512_mic_ONLY > Class Template Reference

Implements a container to dispatch forward layers to cpu-specific implementations. More...

Class Declaration

template<ComputeMode mode, typename sse2Container DAAL_KERNEL_AVX512_mic_ONLY>
class daal::algorithms::neural_networks::layers::forward::interface1::AlgorithmDispatchLayerContainer< mode, DAAL_KERNEL_AVX512_mic_ONLY >

Template Parameters
sse2ContainerImplementation for Intel(R) Streaming SIMD Extensions 2 (Intel(R) SSE2)
ssse3ContainerImplementation for Supplemental Streaming SIMD Extensions 3 (SSSE3)
sse42ContainerImplementation for Intel(R) Streaming SIMD Extensions 42 (Intel(R) SSE42)
avxContainerImplementation for Intel(R) Advanced Vector Extensions (Intel(R) AVX)
avx2ContainerImplementation for Intel(R) Advanced Vector Extensions 2 (Intel(R) AVX2)
avx512_micContainerImplementation for Intel(R) Xeon Phi(TM) processors/coprocessors based on Intel(R) Advanced Vector Extensions 512 (Intel(R) AVX512)
avx512ContainerImplementation for Intel(R) Xeon(R) processors based on Intel AVX-512

Constructor & Destructor Documentation

AlgorithmDispatchLayerContainer ( daal::services::Environment::env *  daalEnv)

Default constructor. Constructs empty container

Member Function Documentation

virtual services::Status allocateInput ( )
inlinevirtual

Allocates weights and biases tensors if they exist

Reimplemented from LayerContainerIfaceImpl.

virtual services::Status initializeInput ( )
inlinevirtual

Initializes values of weights and biases

Reimplemented from LayerContainerIfaceImpl.

virtual services::Status setupCompute ( )
inlinevirtual

Returns a serialization tag, a unique identifier of this class used in serialization

Returns
Serialization tag

Reimplemented from LayerContainerIfaceImpl.


The documentation for this class was generated from the following file:

For more complete information about compiler optimizations, see our Optimization Notice.