Characterizing convolutional neural network workloads on a detailed GPU simulator

Kwanghee Chang, Minsik Kim, Kyungah Kim, Won Woo Ro

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Recent frameworks on convolutional neural networks (CNNs) such as Caffe and MXNet have focused primarily on being compatible with CUDA software and hardware application. However, it was designed for GPU architecture of compute capability 3.0 and above. Therefore, it needs verification of function to perform GPGPU-Sim which is implemented as NVIDIA compute capability devices 2.x. We developed a framework which can make inferencing AlexNet on GPGPU-Sim. We also analyze the execution results of the GPGPU-Sim. The number of lines in one set of the L1 data cache is sensitive to influence performance of AlexNet inference.

Original languageEnglish
Title of host publicationProceedings - International SoC Design Conference 2017, ISOCC 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages84-85
Number of pages2
ISBN (Electronic)9781538622858
DOIs
Publication statusPublished - 2018 May 29
Event14th International SoC Design Conference, ISOCC 2017 - Seoul, Korea, Republic of
Duration: 2017 Nov 52017 Nov 8

Publication series

NameProceedings - International SoC Design Conference 2017, ISOCC 2017

Other

Other14th International SoC Design Conference, ISOCC 2017
CountryKorea, Republic of
CitySeoul
Period17/11/517/11/8

    Fingerprint

All Science Journal Classification (ASJC) codes

  • Hardware and Architecture
  • Electrical and Electronic Engineering
  • Electronic, Optical and Magnetic Materials

Cite this

Chang, K., Kim, M., Kim, K., & Ro, W. W. (2018). Characterizing convolutional neural network workloads on a detailed GPU simulator. In Proceedings - International SoC Design Conference 2017, ISOCC 2017 (pp. 84-85). (Proceedings - International SoC Design Conference 2017, ISOCC 2017). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ISOCC.2017.8368781