Selected Research Works with High "Impact Factors"

Overview

Our research has been recognized for many pioneering research projects on developing high performance computing and data management systems on emerging hardware (particularly on GPUs and FPGAs) and cloud computing.

Unlike citation-centric impact factor measurement for journal publications, we strongly believe that the impact of system research should have more comprehensive measurements to reflect its value to academia and industry. Specifically, we have the following “Impact Factors” for system research: Citations, Relevance to Industry and Open-Source Community, System Repeatability and Academic Impacts, Educational Adoptions, and Media Coverage. (see detailed definition)


  • Bingsheng He, Wenbin Fang, Qiong Luo, Naga K. Govindaraju, Tuyong Wang. Mars: A MapReduce Framework on Graphics Processors. ACM/IEEE International Conference on Parallel Architecture and Compilation Techniques (PACT) 2008. (an extended version of integrating Mars into Hadoop published in IEEE TPDS).

    Our system Mars has inspired the usage of heterogeneous architectures in big data systems such as Hadoop and Spark. Our Mars paper is the 2nd mostly cited among all the papers published in ACM PACT (a leading conference in parallel computing), according to ACM Digital Library (see details)


  • Bingsheng He, Ke Yang, Rui Fang, Mian Lu, Naga K. Govindaraju, Qiong Luo, Pedro V. Sander. Relational Joins on Graphics Processors. ACM SIGMOD International Conference on Management of data, pages: 511-524, 2008. (“Best papers” in ACM SIGMOD 2008, the extension invited to be featured at a special issue at ACM TODS)

    Our system GPUQP is the pioneering system of accelerating databases on GPUs. GPUQP has been the source of inspirations for many GPU-accelerated relational databases in academia and industry players including Brytlyt, BlazingDB, Omnisci (formerly MapD) and SQream. (see details)


  • Jianlong Zhong and Bingsheng He. Medusa: Simplified Graph Processing on GPUs. IEEE Transactions on Parallel and Distributed System, vol.25, no.6, pp.1543-1552, June 2014. (Medusa system demo was among the “Best demos” in VLDB 2013, invited to be featured at "Systems and Prototypes" at ACM SIGMOD RECORD).

    Our system Medusa is the pioneering system of accelerating graph processing on GPUs, which inspired many GPU-based large graph processing systems in academia and industry players such as BlazeGraph. (see details)


  • Zeyi Wen, Jiashuai Shi, Qinbin Li, Bingsheng He, and Jian Chen. ThunderSVM: A Fast SVM Library on GPUs and CPUs. Journal of Machine Learning Research (JMLR) 19 (2018) 1-5.
  • Zeyi Wen, Bingsheng He, Ramamohanarao Kotagiri, Shengliang Lu, Jiashuai Shi. Efficient Gradient Boosted Decision Tree Training on GPUs. IEEE International Parallel & Distributed Processing Symposium (IPDPS) 2018.

    The open-source systems ThunderSVM and ThunderGBM have attracted over 1,600 stars, 5000+ repostings on social media and dozens of adoptions in research publications in two years. (see details)


  • Jianlong Zhong, Bingsheng He. Kernelet: High-Throughput GPU Kernel Execution with Dynamic Slicing and Scheduling. IEEE Transactions on Parallel and Distributed System, vol.25, no.6, pp.1522-1532, June 2014.
  • Mochi Xue, Kun Tian, Yaozu Dong, Jiajun Wang and Zhengwei Qi, Bingsheng He, Haibing Guan. gScale: Scaling up GPU Virtualization with Dynamic Sharing of Graphics Memory Space. USENIX Annual Technical Conference (ATC) 2016.

    Our research has significant practical impacts in GPU virtualizations and scheduling (nowadays an important infrastructure component in cloud computing with GPUs). gScale has been integrated into Intel's Open GPU virtualization platform. (see details)


  • Zeke Wang, Bingsheng He, Wei Zhang, Shunning Jiang. A Performance Analysis Framework for Optimizing OpenCL Applications on FPGAs. IEEE International Symposium on High Performance Computer Architecture (HPCA) 2016.
  • Jieru Zhao, Liang Feng, Wei Zhang, Sharad Sinha, Yun (Eric) Liang, Bingsheng He. COMBA: A Comprehensive Model-Based Analysis Framework for High Level Synthesis of Real Applications. International Conference On Computer Aided Design (ICCAD) 2017. [William J. McCalla Best Paper Award (Front End)]

    The poineering work related to FPGA optimizations have led the rethinking of system optimizations and performance tuning on new-generation FPGAs, and have attracted broad industry interests (e.g., Microsoft (gift grant) and Xilinx research (infrastructure gift)). (see details)


  • You can check out more details about my system research on emerging storage and cloud computing.

  • Acknowledgement

    This page on my influential works was inspired by the page on "Influential Papers" from Prof. Xiaodong Zhang.

    I consider myself very fortunate to work with many wonderful students and collaborators. Without their hardworking and warm-hearted contributions, those works won’t be possible.

    Credits of "details" pages to: Mr. Xinyu Chen, Mr. Hongshi Tan and Mr. Qinbin Li.

    All Rights Reserved to Bingsheng He © 2020