Art

J-GLOBAL ID：202002238363633325   Reference number：20A0661538

分散深層学習を高速化させるFPGA Ring-Allreduceの検討

Author (8)： TANAKA KENJI (NTT 先端集積デバイス研) ,  ARIKAWA YUKI (NTT 先端集積デバイス研) ,  ITO TAKESHI (NTT 先端集積デバイス研) ,  TERADA KAZUHIKO (NTT 先端集積デバイス研) ,  MORITA KAZUTAKA (NTT ソフトウェアイノベーションセ) ,  MIURA SHIKO (NTT ソフトウェアイノベーションセ) ,  TERAMOTO JUNJI (NTT ソフトウェアイノベーションセ) ,  SAKAMOTO TAKESHI (NTT 先端集積デバイス研)
Material： 情報処理学会全国大会講演論文集
Volume： 82nd  Issue： 1  Page： 1.31-1.32  Publication year： Feb. 20, 2020 
JST Material Number： S0731A  Document type： Proceedings
Article type： 原著論文  Country of issue： Japan (JPN)  Language： JAPANESE (JA)

Thesaurus term： *machine learning ,  learning ,  parallel processing ,  distributed processing ,  *message switching ,  disorder/trouble/obstacle ,  speedup ,  experiment ,  *FPGA
Semi thesaurus term： deep learning ,  bottleneck ,  model learning ,  group communication ,  deep learning ,  evaluation experiment

JST classification (2)： Artificial intelligence  (JE08000Z) ,  Digital computer hardwares in general  (JC04010C)

Reference (4)：

• Tal Ben-Num, and Torsten Hoefler: Demystifying Parallel and Distributed Deep Learning: An In-Depth Concurrency Analysis, arXiv:cs.LG/1802.09941, (2018).
• Takuya Akiba, Shuji Suzuki, and Keisuke Fukuda: Extremely Large Minibatch SGD: Training ResNet-50 on ImageNet in 15 Minutes, Deep Learning at Supercomputer Scale (NIPS’17 Workshop), arXiv:cs.DC/1711.04325, (2017).
• Xianyan Jia, Shutao Song, Wei He, Yangzihao Wang, Haidong Rong, Feihu Zhou, Liqiang Xie, Zhenyu Guo, Yuanzhou Yang, Liwei Yu, Tiegang Chen, Guangxiao Hu, Shaohuai Shi, Xiaowen Chu: Highly Scalable Deep Learning Training System with Mixed-Precision: Training ImageNet in Four Minutes, Workshop on Systems for ML and Open Source Software at NeurIPS 2018, arXiv:cs.CV/1807.11205, (2018).
• Chris Ying, Sameer Kumar, Dehao Chen, Tao Wang, Youlong Cheng: Image Classification at Supercomputer Scale, Workshop on Systems for ML and Open Source Software at NeurIPS 2018, arXiv:cs.CV/1811.06992, (2018).

Terms in the title (2)： 深層学習 ,  FPGA