Class Hierarchy

This inheritance list is sorted roughly, but not completely, alphabetically:

[detail level 1234]

C_activation_param_t	Parameters to the activation function hardware
C_SamplingInfo	Simulation sampling information maintained by the Operator and passed to the accelerated kernel
CActivationInfo	Specifies an activation function and relevant parameters
Csmaug::AvgPoolingOp< Backend >	Implements the arithmetic-average-pooling operator
Csmaug::BatchNormOp< Backend >	Implements the batch normalization layer
►Csmaug::BatchNormOp< SmvBackend >
Csmaug::SmvBatchNormOp	SMV backend implementation of batch normalization
Csmaug::ConcatOp< Backend >	Concatenates N Tensors along a specified axis
Csmaug::ConvolutionOp< Backend >	The base class for all 4D spatial convolution operators
►Csmaug::ConvolutionOp< SmvBackend >
Csmaug::SmvConvolutionOp	SMV backend implementation of convolution
Csmaug::TiledTensor::CopyTilesArgs
Csmaug::DataflowGraphWriter	DataflowGraphWriter writes the current network as a dot-graph file to the given ostream
Csmaug::DataOp< Backend >	Exposes a Tensor as its only output
Csmaug::DebugStream	An stream class to consume debug logs
Csmaug::DepthwiseConvolutionOp< Backend >	Implements the depthwise convolution operator
Csmaug::EltwiseAddOp< Backend >	Adds two Tensors elementwise
►Csmaug::EltwiseAddOp< SmvBackend >
Csmaug::SmvEltwiseAddOp	Elementwise addition on SMV
Csmaug::EltwiseMulOp< Backend >	Multiplies two Tensors elementwise
►Csmaug::EltwiseMulOp< SmvBackend >
Csmaug::SmvEltwiseMulOp	Elementwise multiplication on SMV
Csmaug::EluOp< Backend >	Implements the exponential linear unit function
►Csmaug::EluOp< SmvBackend >
Csmaug::SmvEluOp	Elementwise exponential linear unit on SMV
Csmaug::FlattenOp< Backend >	Flattens each batch of a Tensor
Csmaug::FromDataType< DataType >	Provides compile-time conversion from SMAUG DataType to C type
Csmaug::FromDataType< Bool >
Csmaug::FromDataType< Float16 >
Csmaug::FromDataType< Float32 >
Csmaug::FromDataType< Float64 >
Csmaug::FromDataType< Int32 >
Csmaug::FromDataType< Int64 >
Csmaug::GreaterEqualOp< Backend >	Implements an elementwise greater than or equal to operator
►Csmaug::GreaterEqualOp< SmvBackend >
Csmaug::SmvGreaterEqualOp	Elementwise greater-than-or-equal-to operator on SMV
Csmaug::GreaterOp< Backend >	Implements an elementwise greater than operator
►Csmaug::GreaterOp< SmvBackend >
Csmaug::SmvGreaterOp	Elementwise greater-than operator on SMV
Csmaug::HardTanhOp< Backend >	Implements the hard tanh operator, which bounds the min and max value of the tanh operator
►Csmaug::HardTanhOp< SmvBackend >
Csmaug::SmvHardTanhOp	Hard tanh operator on SMV
Csmaug::InnerProductOp< Backend >	Implements the inner product operator
►Csmaug::InnerProductOp< SmvBackend >
Csmaug::SmvInnerProductOp	Inner product operator on SMV
Csmaug::LessEqualOp< Backend >	Implements an elementwise less-than-or-equal-to operator
►Csmaug::LessEqualOp< SmvBackend >
Csmaug::SmvLessEqualOp	Elementwise less-than-or-equal-to operator on SMV
Csmaug::LessOp< Backend >	Implements an elementwise less-than operator
►Csmaug::LessOp< SmvBackend >
Csmaug::SmvLessOp	Elementwise less-than operator on SMV
Csmaug::MaxPoolingOp< Backend >	Implements the max-pooling operator
Csmaug::MergeOp< Backend >	Forwards the first live input to its output
Csmaug::Network	Network encapsulates all of the information SMAUG will use during execution: the overall computation graph of the model, all the operators and tensors, various housekeeping structures, and simulation information
►Csmaug::Operator	Operator is the base class for all graph operators supported by SMAUG
►Csmaug::PoolingOp< SmvBackend >
►Csmaug::SmvPoolingOp	Base class for SMV pooling oeprators
Csmaug::SmvAvgPoolingOp	Average pooling operator on SMV
Csmaug::SmvMaxPoolingOp	Max-pooling operator on SMV
Csmaug::EltwiseOp< Backend >	The base class of all elementwise operators
Csmaug::FusedActivationOp	An Operator fused with an activation function
Csmaug::PoolingOp< Backend >	Implements a pooling operator
Csmaug::UnaryOp< Backend >	Base class for all operators with one input
Csmaug::Network::OperatorInsertion
Csmaug::PaddingOp< Backend >	Pad a given tensor in any number of dimensions with arbitrary size
Csmaug::ReferenceBackend	ReferenceBackend provides reference implementations of all operators supported by SMAUG
Csmaug::ReluOp< Backend >	Implements the rectified linear unit operator: max(slope * x, 0)
►Csmaug::ReluOp< SmvBackend >
Csmaug::SmvReluOp	Rectified linear-unit operator on SMV
Csmaug::ReorderOp< Backend >	Implements a Tensor reordering operation to convert between different DataLayouts
Csmaug::RepeatOp< Backend >	Replicates a Tensor's data among all dimensions
Csmaug::ReshapeOp< Backend >	Changes the Tensor's shape while retaining the number of elements
Csmaug::Scheduler	Scheduler is responsible for running the Network
Csmaug::gem5::ScopedStats	A RAII helper class which dumps and/or resets gem5 stats at construction and destruction
Csmaug::SeluOp< Backend >	Implements the scaled exponential linear unit function
►Csmaug::SeluOp< SmvBackend >
Csmaug::SmvSeluOp	Elementwise scaled exponential linear unit on SMV
Csmaug::SigmoidOp< Backend >	Implements the sigmoid operator, defined as 1/(1 + exp(-input))
►Csmaug::SigmoidOp< SmvBackend >
Csmaug::SmvSigmoidOp	Sigmoid linear-unit operator on SMV
Csmaug::SmaugTest	The Catch2 test fixture used by all C++ unit tests
Csmaug::SmvAcceleratorPool	Implements a pool of worker accelerators
Csmaug::SmvBackend	SmvBackend implements a set of models of optimized DL kernels that were taped out on a machine learning SoC by the Harvard Architecture, Circuits, and Compilers
Csmaug::SoftmaxOp< Backend >	Implements the softmax operator
►Csmaug::SoftmaxOp< SmvBackend >
Csmaug::SmvSoftmaxOp	Softmax operator on SMV
Csmaug::SplitOp< Backend >	Implements the split operator, which divides a Tensor into N output Tensors along a specified dimension
Csmaug::SwitchOp< Backend >	Conditionally forwards an input to one of two outputs
Csmaug::TanhOp< Backend >	Implements the tanh operator
►Csmaug::TanhOp< SmvBackend >
Csmaug::SmvTanhOp	Tanh operator on SMV
►Csmaug::TensorBase	The base class of all Tensor objects
Csmaug::Tensor	Tensor represents a single multi-dimensional array of data
Csmaug::TiledTensor	A multidimensional container of Tensors
►Csmaug::TensorIndexIterator	An iterator over a multidimensional tensor's indices, accounting for data alignment padding
Csmaug::TensorRegionIndexIterator	A tensor index iterator that stays within a specified rectangular region
Csmaug::TensorIndices	Additional metadata for edges in the graph
Csmaug::TensorShape	TensorShape describes the shape of a Tensor
Csmaug::ThreadPool::ThreadInitArgs
Csmaug::ThreadPool	A user-space cooperatve thread pool implementation designed for gem5 in SE mode
Csmaug::TiledTensor::Tile	A tile is a rectangular portion of a larger Tensor
Csmaug::smv::TilingConfig	A TilingConfig describes tiling strategies and optimal tile sizes for inputs, weights, and outputs Tensors
►Csmaug::smv::TilingOptimizerBase
Csmaug::smv::bn::TilingOptimizer	Tiling optimizer for SMV batch norm kernel
Csmaug::smv::conv::TilingOptimizer	Tiling optimizer for SMV convolution kernel
Csmaug::smv::fc::TilingOptimizer	Tiling optimizer for SMV inner product kernel
Csmaug::smv::pool::TilingOptimizer	Tiling optimizer for pooling operators on SMV
Csmaug::ToDataType< T >	Provides compile-time conversion from C types to SMAUG DataTypes
Csmaug::ToDataType< bool >
Csmaug::ToDataType< double >
Csmaug::ToDataType< float >
Csmaug::ToDataType< float16 >
Csmaug::ToDataType< int32_t >
Csmaug::ToDataType< int64_t >
Csmaug::ToDataType< uint32_t >
Csmaug::ToDataType< uint64_t >
Csmaug::ThreadPool::WorkerThread	All state and metadata for a worker thread
Csmaug::Workspace	Workspace is the container and owner of all Tensors and Operators in the Network