Tron Logo
Trade on HitBTC

Bingsheng He

Associate Professor

About

As an enthusiastic person with a strong background and expertise in designing, implementing and optimizing parallel and distributed systems, I would like to devote myself to system research. After obtaining my Ph.D. in August 2008, I joined the System Research Group of Microsoft Research Asia. My focus was on improving the large-scale data analysis platform for Bing Search. I was a research follow in CUHK for half one year (Feb 2010-July 2010), with a focus on high-throughput transaction processing systems on many-core processors. I joined School of Computer Engineering at Nanyang Technological University as a tenure-track assistant professor in August 2010. I have been growing into a multidisciplinary researcher with exposure to different cultures, research domains and collaborations. I have had a track record of research, since I started my Ph.D. study. I have actively published research results in recognized conferences such as SuperComputing, SIGMOD, PVLDB/VLDB, PACT, CIDR and ICDE, and journals including IEEE TPDS, IEEE TKDE and ACM TODS. I have been dedicating myself to pushing the state-of-the-art in building faster, greener and more scalable parallel and distributed systems using emerging hardware and software technologies. The system design and implementation are often driven by emerging applications.

Recently, we have made some systems open-sourced:
* GPUQP: Query Co-Processing Using Graphics Processors
http://www.cse.ust.hk/gpuqp/
[the pioneering database accelerated by database, which influence many later research and projects in industry.]

* ThunderSVM: A Fast SVM Library on GPUs and CPUs
https://github.com/zeyiwen/thundersvm
[500+ stars by Dec 31, 2017. highlighted by hacker news on Dec 28 2017, https://twitter.com/newsycombinator]

* Relational query processing (in-memory databases) on many-core CPUs.
https://github.com/PatrickXC/hash_join_codes_KNL
[bare metal and vectorized implementations of hash join algorithms on Intel Xeon Phi (KNL)]

Cookies help us deliver our services. By using our services, you agree to our use of cookies.

Learn more Got it