Apache Arrow, a new open source project, is designed to provide high performance in-memory analytics across different systems. It eliminates cross system communication overhead cost and can improve performance 100 times. It is based on the code of Apache Drill. Developers from other Apache projects. It removes the burden of serialization and deserialization of data. It also supports complex data in addition to relational data. Read the complete article here: http://www.cio.com/article/3034279/big-data-gets-a-new-open-source-project-apache-arrow.html