Get MapReduce jobs done faster with half the hardware

MapR’s advanced engineering and architectural innovations get MapReduce jobs done faster with less hardware. In fact, with MapR your jobs will run faster with half the hardware required by other Apache Hadoop distributions. For example, your Hadoop application will run faster on a 50-node MapR cluster than on a 100-node cluster running any other Hadoop distribution.

	Lockless Storage Services™ eliminates contention Direct block device I/O provides throughput at hardware speed Optimized Hadoop Direct Shuffle™ algorithm exploits No NameNode HA™ scalability Automatic & transparent client-side compression reduces network overhead and reduces footprint on disk System written in C++ eliminating sporadic Java garbage collection overhead