HAMR is a natural and intuitive evolution of Hadoop programming paradigms, providing a unified feature-rich framework that is fast, easy to use, and expressive

HAMR’s programming model is recognizable by Hadoop developers

HAMR’s runtime utilizes its patent-pending Flowlets™ technology to provide performance up to 20X vs. standard Hadoop

HAMR runs within Hadoop ecosystem (maintains the best aspects of MapReduce in Apache Hadoop 2.0) as well as stand-alone