登录 注册 会员服务 联系我们

NIOT: A Novel Inference Optimization of Transformers on Modern CPUs

  • Published:2023-06 Issue:6 Volume:34 Page:1982-1995
  • ISSN:1045-9219
  • Container-title:IEEE Transactions on Parallel and Distributed Systems
  • language:
  • Short-container-title:IEEE Trans. Parallel Distrib. Syst.

Author:

Zhang Zining1ORCID,Chen Yao1ORCID,He Bingsheng1ORCID,Zhang Zhenjie2

Affiliation:

1. School of Computing, National University of Singapore, Singapore

2. Neuron Mobility Pte. Ltd., Singapore

Funder

National Research Foundation Singapore

AI Singapore Programme

Ministry of Education AcRF Tier 2

Centre for Trusted Internet and Community

National University of Singapore

Publisher

Institute of Electrical and Electronics Engineers (IEEE)

Subject

Computational Theory and Mathematics,Hardware and Architecture,Signal Processing

Reference58 articles.

1. Efficient 8-bit quantization of transformer neural machine language translation model;bhandare,2019

2. Tacker: Tensor-CUDA Core Kernel Fusion for Improving the GPU Utilization while Ensuring QoS

3. Losing Heads in the Lottery: Pruning Transformer Attention in Neural Machine Translation

4. TransMask: A Compact and Fast Speech Separation Model Based on Transformer

5. Generating long sequences with sparse transformers;child,2019

Cited by 3 articles. 订阅此论文施引文献 订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献

1. O-RAN-Based Digital Twin Function Virtualization for Sustainable IoV Service Response: An Asynchronous Hierarchical Reinforcement Learning Approach;IEEE Transactions on Green Communications and Networking;2024-09

2. Optimizing Attention by Exploiting Data Reuse on ARM Multi-core CPUs;Proceedings of the 38th ACM International Conference on Supercomputing;2024-05-30

3. YFlows: Systematic Dataflow Exploration and Code Generation for Efficient Neural Network Inference using SIMD Architectures on CPUs;Proceedings of the 33rd ACM SIGPLAN International Conference on Compiler Construction;2024-02-17
同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线，采集、加工和组织学术论文而形成的新型学术文献查询和分析系统，可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容，当前同舟云学术共收录了国内外主流学术期刊6万余种，收集的期刊论文及会议论文总量共计约1.5亿篇，并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询！咨询电话：010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3