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.
Classes | Variables
zigzag.workload.layer_attributes Namespace Reference

Classes

class  LayerEquation
 "! core computation equation, e.g. More...
 
class  LayerDimSizes
 Contains the size of each computation loop as defined in the workload, e.g. More...
 
class  LayerOperandPrecision
 Contains the bit precision of each layer operand. More...
 
class  MemoryOperandLinks
 Links LayerOperand to MemoryOperand. More...
 
class  LayerDimRelation
 For the operand dimension that is not directly a loop dimension, a relation equations between them(operand dimension) and the loop dimension is required. More...
 
class  LayerTemporalOrdering
 Represents a user-defined temporal ordering. More...
 
class  LayerPadding
 

Variables

 logger = logging.getLogger(__name__)
 
 TypeAlias
 

Variable Documentation

◆ logger

logger = logging.getLogger(__name__)

◆ TypeAlias

TypeAlias