A MAC-less Neural Inference Processor Supporting Compressed, Variable Precision Weights

This paper introduces two architectures for the inference of convolutional neural networks (CNNs). Both architectures exploit weight sparsity and compression to reduce computational complexity and bandwidth. The first architecture uses multiply-accumulators (MACs) but avoids unnecessary multiplicati...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	arXiv.org 2020-12
1. Verfasser:	Liguori, Vincenzo
Format:	Artikel
Sprache:	eng
Schlagworte:	Accumulators Arithmetic coding Artificial neural networks Inference Microprocessors Sparsity Weight
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!