UniSched: A Unified Scheduler for Deep Learning Training Jobs With Different User Demands-Reference-Cited by-同舟云学术

登录注册会员服务联系我们

UniSched: A Unified Scheduler for Deep Learning Training Jobs With Different User Demands

Published:2024-06 Issue:6 Volume:73 Page:1500-1515
ISSN:0018-9340
Container-title:IEEE Transactions on Computers
language:
Short-container-title:IEEE Trans. Comput.

Author:

Gao Wei¹^ORCID,Ye Zhisheng²^ORCID,Sun Peng³^ORCID,Zhang Tianwei¹^ORCID,Wen Yonggang¹^ORCID

Affiliation:

1. School of Computer Science and Engineering, Nanyang Technological University, Singapore

2. Peking University, Beijing, China

3. Shanghai AI Lab & SenseTime, Beijing, China

Funder

National Key R&D Program of China

RIE2020 Industry Alignment Fund - Industry Collaboration Projects

Publisher

Institute of Electrical and Electronics Engineers (IEEE)

Link

http://xplorestaging.ieee.org/ielx7/12/10527219/10454114.pdf?arnumber=10454114

Reference40 articles.

1. Gandiva: Introspective cluster scheduling for deep learning;Xiao,2018

2. Analysis of large-scale multi-tenant GPU clusters for DNN training workloads;Jeon,2019

3. Tiresias: A GPU cluster manager for distributed deep learning;Gu,2019

4. HiveD: Sharing a GPU cluster for deep learning with guarantees;Zhao,2020

5. AntMan: Dynamic scaling on GPU clusters for deep learning;Xiao,2020

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