PRAP-PIM: A weight pattern reusing aware pruning method for ReRAM-based PIM DNN accelerators

Resistive Random-Access Memory (ReRAM) based Processing-in-Memory (PIM) frameworks are proposed to accelerate the working process of DNN models by eliminating the data movement between the computing and memory units. To further mitigate the space and energy consumption, DNN model weight sparsity and...

Full description

Bibliographic Details

Published in:	High-Confidence Computing
Main Authors:	Zhaoyan Shen, Jinhao Wu, Xikun Jiang, Yuhao Zhang, Lei Ju, Zhiping Jia
Format:	Article
Language:	English
Published:	Elsevier 2023-06-01
Subjects:	Resistive Random-Access Memory Processing-in-Memory Deep neural network Model compression
Online Access:	http://www.sciencedirect.com/science/article/pii/S2667295223000211