Ongoing projects
Benchmark proposals.
Formation of working groups.
Publish benchmark specifications.
Open-source implementations of benchmark specifications.
Challenges and competitions.
Archive performance numbers.
Eight new proposals are going to be present at Bench 18.
Big Data Benchmarking: Applications and Systems(Slides)
Prof. Geoffrey Fox, Indiana University
MLPerf: The Vision Behind an ML Benchmark Suite for Measuring the Performance of ML Software Frameworks, ML Hardware Accelerators, and ML Cloud and Edge Platforms(Slides)
Prof. Vijay Janapa Reddi, Harvard University
DataMotif: A Benchmark Proposal for Big Data and AI(Slides)
Dr. Wanling Gao, ICT, CAS
A Benchmark proposal for Deep Learning Benchmarks(Slides)
Prof. Xiaoyi Lu, The Ohio State University
A Benchmark proposal for Datacenter Computing(Slides)
Dr. Chen Zheng, ICT, CAS
PeakBench: A Benchmark Proposal for Scalable Transaction Processing(Slides)
Prof. Weining Qian, East China Normal University
TS-benchmark: a benchmark proposal for time series databases(Slides)
Prof. Yueguo Chen, Renmin University of China
A Benchmark proposal for large-scale and high-speed spatiotemporal data processing and analytic(Slides)
Prof. Zhiyuan Chen, Prof. Jianwu Wang, Univiersity of Maryland, Baltimore County
Available Soon
TBD
TBD
Will be organized at Bench 19 and BenchCouncil 2019 Annual system technology conference.
TBD
Solicit Benchmark Proposals
Benchmark proposals are solicited on any topic, including, but not limited to:
Artificial Intelligence
NewSQL and distributed database systems
Blockchain
Datacenter, cloud and warehouse-scale computing
High performance computing
Mobile Robotics
Edge and fog computing
Big Scientific Data
Internet of Things (IoT)
Education, financial and power system
According to BenchCouncil’s procedure, each benchmarking pipeline consists of the six steps: benchmark proposal, formation of working group, publish benchmark specifications, open-source implementations of benchmark specifications, challenges and competitions, and archive performance numbers.