Mosharaf Chowdhury

Mosharaf Chowdhury
Alma materUniversity of California, Berkeley, Bangladesh University of Engineering and Technology
Awards
Scientific career
InstitutionsUniversity of Michigan, Ann Arbor
ThesisCoflow: A Networking Abstraction for Distributed Data-Parallel Applications (2015)
Doctoral advisorIon Stoica
Websitewww.mosharaf.com

Mosharaf Chowdhury is a Bangladeshi-American computer scientist known for his contributions to the fields of computer networking and large-scale systems for emerging machine learning and big data workloads. He is an Associate Professor of Computer Science and Engineering at the University of Michigan, Ann Arbor and leads SymbioticLab. He is the creator of coflow[2] and the co-creator of Apache Spark.[3]

Research

Chowdhury specializes in the fields of computer networking and large-scale systems for emerging machine learning and big data workloads. Especially, his research aims for the symbiosis of AI/ML applications and software/hardware infrastructure in wide-area, datacenter-scale, and rack-scale computing.

Chowdhury pioneered many fields of research and technology in the context of emerging workloads and computer systems. Chowdhury created Infiniswap, the first practical memory disaggregation system,[4] Salus, the first software-only GPU sharing system for deep learning,[5] FedScale, the largest federated learning benchmark and platform,[6] and Zeus, the first GPU time & energy optimization framework for deep learning.[7]

References

  1. ^ "SIGCOMM Doctoral Dissertation Award | acm sigcomm". www.sigcomm.org. Retrieved 2023-05-07.
  2. ^ Chowdhury, Mosharaf; Stoica, Ion (2012-10-29). "Coflow: A networking abstraction for cluster applications". Proceedings of the 11th ACM Workshop on Hot Topics in Networks. HotNets-XI. New York, NY, USA: Association for Computing Machinery. pp. 31–36. doi:10.1145/2390231.2390237. ISBN 978-1-4503-1776-4. S2CID 6956491.
  3. ^ Zaharia, Matei; Chowdhury, Mosharaf; Franklin, Michael J.; Shenker, Scott; Stoica, Ion (2010-06-22). "Spark: cluster computing with working sets". Proceedings of the 2nd USENIX Conference on Hot Topics in Cloud Computing. HotCloud'10. USA: USENIX Association.
  4. ^ Gu, Juncheng; Lee, Youngmoon; Zhang, Yiwen; Chowdhury, Mosharaf; Shin, Kang G. (2017). "Efficient Memory Disaggregation with Infiniswap". Proceedings of the 14th USENIX Symposium on Networked Systems Design and Implementation (NSDI ’17). pp. 649–667. ISBN 978-1-931971-37-9.
  5. ^ Yu, Peifeng; Chowdhury, Mosharaf (2020). "Fine-Grained GPU Sharing Primitives for Deep Learning Applications" (PDF). Proceedings of the 3rd MLSys Conference.
  6. ^ Lai, Fan; Dai, Yinwei; Singapuram, Sanjay; Liu, Jiachen; Zhu, Xiangfeng; Madhyastha, Harsha; Chowdhury, Mosharaf (2022-06-28). "FedScale: Benchmarking Model and System Performance of Federated Learning at Scale". Proceedings of the 39th International Conference on Machine Learning. PMLR: 11814–11827. arXiv:2105.11367.
  7. ^ You, Jie; Chung, Jae-Won; Chowdhury, Mosharaf (2023). "Zeus: Understanding and Optimizing {GPU} Energy Consumption of {DNN} Training". Usenix Nsdi: 119–139. ISBN 978-1-939133-33-5.