A Weight Mapping Strategy for More Fully Exploiting Data in CIM-Based CNN Accelerator

Shang Wang,Feng Liang,Qi Cao, Yongqiang Wang,Haoyuan Li,Junzhe Liang

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-EXPRESS BRIEFS（2024）

Xi An Jiao Tong Univ | Simon Fraser Univ

Cited 0|Views8

Abstract

Compute-in-memory accelerators have been extensively researched to overcome the limitations of the von Neumann architecture. However, the current mapping strategy and dataflow results in inefficient utilization of the array and input data. In this paper, we propose a new mapping method named Squeezemapping that leverages spare space in each array and optimizes the utilization of input dataset. We employed NeuroSim to simulate the inference of various networks of different scales. Experimental results demonstrate that our method performs 36.51% higher in energy efficiency and 48.15% higher in speedup when applied to the VGG16 large-scale model under area constraints.

Translated text

Key words

Kernel,Information filters,Convolutional neural networks,Computer architecture,Arrays,Convolution,Energy efficiency,Convolutional neural network,weight mapping method,compute-in-memory,data reuse,array utilization

求助PDF

上传PDF

Bibtex

AI Read Science

AI Summary

AI Summary is the key point extracted automatically understanding the full text of the paper, including the background, methods, results, conclusions, icons and other key content, so that you can get the outline of the paper at a glance.

Example

Background

Key content

Introduction

Methods

Results

Related work

Fund

Key content

Pretraining has recently greatly promoted the development of natural language processing (NLP)
We show that M6 outperforms the baselines in multimodal downstream tasks, and the large M6 with 10 parameters can reach a better performance
We propose a method called M6 that is able to process information of multiple modalities and perform both single-modal and cross-modal understanding and generation
The model is scaled to large model with 10 billion parameters with sophisticated deployment, and the 10 -parameter M6-large is the largest pretrained model in Chinese
Experimental results show that our proposed M6 outperforms the baseline in a number of downstream tasks concerning both single modality and multiple modalities We will continue the pretraining of extremely large models by increasing data to explore the limit of its performance

Upload PDF to Generate Summary

Must-Reading Tree

Example

Generate MRT to find the research sequence of this paper