ZigZag - Deep Learning Hardware Design Space Exploration
This repository presents the novel version of our tried-and-tested hardware Architecture-Mapping Design Space Exploration (DSE) Framework for Deep Learning (DL) accelerators. ZigZag bridges the gap between algorithmic DL decisions and their acceleration cost on specialized accelerators through a fast and accurate hardware cost estimation.
Static Public Attributes | List of all members
Constants Class Reference

Store constant objects used throughout ZigZag (instead of hardcoding them) More...

Static Public Attributes

 LAYER_OP_I = LayerOperand("I")
 
 LAYER_OP_W = LayerOperand("W")
 
 OUTPUT_LAYER_OP = LayerOperand("O")
 
 FINAL_OUTPUT_LAYER_OP = LayerOperand("O_final")
 
 MEM_OP_1 = MemoryOperand("I1")
 
 MEM_OP_2 = MemoryOperand("I2")
 
 OUTPUT_MEM_OP = MemoryOperand("O")
 
 FINAL_OUTPUT_MEM_OP = MemoryOperand("O_final")
 
 UNKNOWN_DIM_OPERATOR = LayerDim("*")
 
 TypeAlias
 Type aliases ######. More...
 

Detailed Description

Store constant objects used throughout ZigZag (instead of hardcoding them)

Member Data Documentation

◆ FINAL_OUTPUT_LAYER_OP

FINAL_OUTPUT_LAYER_OP = LayerOperand("O_final")
static

◆ FINAL_OUTPUT_MEM_OP

FINAL_OUTPUT_MEM_OP = MemoryOperand("O_final")
static

◆ LAYER_OP_I

LAYER_OP_I = LayerOperand("I")
static

◆ LAYER_OP_W

LAYER_OP_W = LayerOperand("W")
static

◆ MEM_OP_1

MEM_OP_1 = MemoryOperand("I1")
static

◆ MEM_OP_2

MEM_OP_2 = MemoryOperand("I2")
static

◆ OUTPUT_LAYER_OP

OUTPUT_LAYER_OP = LayerOperand("O")
static

◆ OUTPUT_MEM_OP

OUTPUT_MEM_OP = MemoryOperand("O")
static

◆ TypeAlias

TypeAlias
static

Type aliases ######.

◆ UNKNOWN_DIM_OPERATOR

UNKNOWN_DIM_OPERATOR = LayerDim("*")
static
The documentation for this class was generated from the following file: