咨询与建议

限定检索结果

文献类型

  • 1 篇 期刊文献

馆藏范围

  • 1 篇 电子文献
  • 0 种 纸本馆藏

日期分布

学科分类号

  • 1 篇 工学
    • 1 篇 电气工程
    • 1 篇 计算机科学与技术...
    • 1 篇 软件工程

主题

  • 1 篇 cache access lat...
  • 1 篇 cache storage
  • 1 篇 neural chips
  • 1 篇 deep learning
  • 1 篇 gpu fpga heterog...
  • 1 篇 field programmab...
  • 1 篇 training data
  • 1 篇 training strateg...
  • 1 篇 image net
  • 1 篇 high bandwidth i...
  • 1 篇 graphical proces...
  • 1 篇 bandwidth
  • 1 篇 random access me...
  • 1 篇 data aggregation
  • 1 篇 collective commu...
  • 1 篇 deep learning ar...
  • 1 篇 res net 50
  • 1 篇 novel allreduce ...
  • 1 篇 allreduce latenc...
  • 1 篇 network interfac...

机构

  • 1 篇 ntt corp ntt dev...
  • 1 篇 ntt corp ntt sof...

作者

  • 1 篇 morita kazutaka
  • 1 篇 tanaka kenji
  • 1 篇 ito tsuyoshi
  • 1 篇 nemoto naru
  • 1 篇 terada kazuhiko
  • 1 篇 sakamoto takeshi
  • 1 篇 teramoto junji
  • 1 篇 arikawa yuki

语言

  • 1 篇 英文
检索条件"主题词=GPU FPGA Heterogeneous Computing"
1 条 记 录,以下是1-10 订阅
排序:
Distributed Deep Learning With gpu-fpga heterogeneous computing
收藏 引用
IEEE MICRO 2021年 第1期41卷 15-22页
作者: Tanaka, Kenji Arikawa, Yuki Ito, Tsuyoshi Morita, Kazutaka Nemoto, Naru Terada, Kazuhiko Teramoto, Junji Sakamoto, Takeshi NTT Corp NTT Device Technol Labs Atsugi Kanagawa 2430198 Japan NTT Corp NTT Software Innovat Ctr Tokyo 1808585 Japan
In distributed deep learning (DL), collective communication algorithms, such as Allreduce, used to share training results between graphical processing units (gpus) are an inevitable bottleneck. We hypothesize that a c... 详细信息
来源: 评论