ElasticFlow: An Elastic Serverless Training Platform for Distributed Deep Learning

Author:

Gu Diandian1,Zhao Yihao1,Zhong Yinmin1,Xiong Yifan2,Han Zhenhua2,Cheng Peng2,Yang Fan2,Huang Gang1,Jin Xin1,Liu Xuanzhe1

Affiliation:

1. Peking University, China

2. Microsoft Research, China

Funder

National Natural Science Foundation of China

National Natural Science Fund for the Excellent Young Scientists Fund Program (Overseas)

Beijing Outstanding Young Scientist Program

Microsoft University Collaboration Program

PKU-Baidu Fund Project

Publisher

ACM

Reference70 articles.

1. 2019. NCCL. https://developer.nvidia.com/nccl Retrieved on July 3, 2022 2019. NCCL. https://developer.nvidia.com/nccl Retrieved on July 3, 2022

2. 2021. AWS EC2 pricing. https://aws.amazon.com/ec2/pricing Retrieved on December 23, 2022 2021. AWS EC2 pricing. https://aws.amazon.com/ec2/pricing Retrieved on December 23, 2022

3. 2022. Amazon SageMaker . https://aws.amazon.com/sagemaker/?nc1=h_ls Retrieved on July 3, 2022 2022. Amazon SageMaker. https://aws.amazon.com/sagemaker/?nc1=h_ls Retrieved on July 3, 2022

4. 2022. ElasticFlow Traces . https://github.com/microsoft/elasticflow-traces Retrieved on December 22, 2022 2022. ElasticFlow Traces. https://github.com/microsoft/elasticflow-traces Retrieved on December 22, 2022

5. 2022. gRPC. https://grpc.io Retrieved on July 3, 2022 2022. gRPC. https://grpc.io Retrieved on July 3, 2022

Cited by 6 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. Auto-scaling mechanisms in serverless computing: A comprehensive review;Computer Science Review;2024-08

2. QUART: Latency-Aware FaaS System for Pipelining Large Model Inference;2024 IEEE 44th International Conference on Distributed Computing Systems (ICDCS);2024-07-23

3. Paldia: Enabling SLO-Compliant and Cost-Effective Serverless Computing on Heterogeneous Hardware;2024 IEEE International Parallel and Distributed Processing Symposium (IPDPS);2024-05-27

4. Cross-System Analysis of Job Characterization and Scheduling in Large-Scale Computing Clusters;2024 IEEE International Parallel and Distributed Processing Symposium (IPDPS);2024-05-27

5. Deep Learning Workload Scheduling in GPU Datacenters: A Survey;ACM Computing Surveys;2024-01-22

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