Deep Fusion: A Software Scheduling Method for Memory Access Optimization - Network and Parallel Computing Access content directly
Conference Papers Year : 2019

Deep Fusion: A Software Scheduling Method for Memory Access Optimization

Abstract

Deep neural networks (DNNs) have been considered to be the state-of-the-art artificial intelligence methods in a very broad range of applications. However, DNNs are compute intensive and memory intensive which are difficult to be employed in practical scenarios. Due to their favorable parallel computing ability, a series of DNN accelerators have been proposed. However, the improvement of on-chip computing capacity and the increasing number of parameters in the neural networks make access to memory a bottleneck. In this paper, we analyze the existing DNN algorithms. We observe that the special structure of neural networks makes it have two useful characteristics, which are unilateral directivity and local independence. Based on these characteristics, we propose a general software scheduling method to reduce memory access cost. Based on the experimental results, our method can reduce 32% memory access cost and achieve a speedup of 1.6x in average on our experiment platform and the best result is in ResNet-50, which is up to 56% and 2.62x.
Fichier principal
Vignette du fichier
486810_1_En_22_Chapter.pdf (1.82 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03770545 , version 1 (06-09-2022)

Licence

Attribution

Identifiers

Cite

Yimin Zhuang, Shaohui Peng, Xiaobing Chen, Shengyuan Zhou, Tian Zhi, et al.. Deep Fusion: A Software Scheduling Method for Memory Access Optimization. 16th IFIP International Conference on Network and Parallel Computing (NPC), Aug 2019, Hohhot, China. pp.277-288, ⟨10.1007/978-3-030-30709-7_22⟩. ⟨hal-03770545⟩
16 View
32 Download

Altmetric

Share

Gmail Facebook X LinkedIn More