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 | |
class | SearchInterLayerDataLocalityStage |
Class for searching lowest allowed memory level per operand per layer. More... | |
class | ExploitInterLayerDataLocalityStage |
This stage must be processed behind WorkloadStage. More... | |
Namespaces | |
zigzag.stages.exploit_data_locality_stages | |
ZigZag simulates a workload layer-by-layer. | |
Variables | |
logger = logging.getLogger(__name__) | |