• The domain-specific big data platform

    BAFST in collaboration with China Academy of Telecommunication Research, and other industrial leading enterprises establish industry-standard big data benchmarks in different areas, e.g., finance and telecommunications. BAFST participates in the development and maintenance of an open source cutting-edge big data benchmarking project--BigDataBench( English, Chinese ). BAFST releases the fist third-party report on big data systems ranking in the world: (BigData100 ).

  • The open platform for accelerating SMEs

    BAFST explores how to use Internet, big data, and modern finance techniques to accelerate the development of SMEs (Small and Medium-sized Enterprises).

  • The open platform for sharing economy

    BAFST explores how to use Internet, big data, and modern finance techniques to set up an open platform for sharing economy, and finally increases the utilization of social resources.

  • BenchIC: Cloud-base benchmarking and evaluation systems

    Benchmarking is the foundation for quantitatively evaluating computer systems. However, benchmarking is hard and requires a lot of time for deployment, installation, debugging, data collection, and analysis. Benchmarking in cloud platforms can reduce the complexity of benchmarking. We build a system named BenchIC (Benchmark in Cloud) to transfer the benchmarking work into the cloud, which enables users only focus on the result analysis while the other routine work automatically done by BenchIC. Currently, BenchIC supports Hadoop, Spark, Hive, SparkSql, Impala, Storm, Spark Streaming, and most NoSQL systems. It also provides the big data generation tools to generate TB to PB real data sets.

  • IOT and Big Data Emulation Platform

    IOT and Big Data Emulation Platform is a test bed for IOT and Big Data. The Emulation Platform can emulate variety of events and properties under different IOT and Big Data scenarios, and collect user-defined data for analysis. The generated data of the Emulation Platform is from multidimensional closer to real-world data. Using the Emulation Platform, users can do any emulation tests without the hardware cost constraints. On the other hand, physical sensor data can be plugged into the Emulation Platform seamlessly.

Copyright © 2015 Beijing Academy of Frontier Science & Technology, Inc.