Intel Software Adrenaline

Optimizing Java* and Apache Hadoop* for Intel® Architecture

Issue link:

Contents of this Issue


Page 0 of 4

WHITE PAPER Intel® Xeon® Processors Optimizing Java* and Apache Hadoop* Optimizing Java* and Apache Hadoop* for Intel® Architecture With the ability to analyze virtually unlimited amounts of unstructured and semistructured data, Apache Hadoop* has the potential to transform data analytics. Yet with Apache Hadoop only recently becoming a more mainstream technology, benchmarking and deployment tools are still catching up, which can make realizing the full potential of a performance-optimized Apache Hadoop cluster challenging. As with most new technologies, Apache Hadoop has gone from being an interesting concept to an established technology. And as new technologies become more established, they become easier to optimize. For example, rail companies, whose technologies and processes are well known, realize the value of optimization. From the performance of the locomotive engines, to the routes that mile-long trains take, optimization is a key factor in lowering the cost per pound of cargo and getting that cargo to its destination efficiently. The same principle applies to choosing the right computing platform that is optimized to squeeze the last bit of performance out of Apache Hadoop. A non-optimized Apache Hadoop cluster might still get the job done, but data analytics tasks will take longer and be less efficient. Systems administrators can increase Apache Hadoop cluster performance by using existing Intel technologies, such as Intel® Solid State Drives (SSDs) and Intel® Ethernet 10 Gigabit Server Adapters. These technologies, combined with Intel® Xeon® processors, can improve a one terabyte (TB) MapReduce sort from four hours to seven minutes.1 Yet because Apache Hadoop is built on Java*, one of the most effective ways to increase performance is to optimize Java itself to take advantage of Intel architecture enhancements. Intel and Oracle: Building a Better Java Foundation Through a collaborative effort between Intel and Oracle software engineers, organizations can realize significant performance gains using the latest Oracle Java Virtual Machine (JVM) running on servers powered by Intel Xeon processors. Intel and Oracle have worked together for the past six years to improve the overall performance of Java running on the Intel Xeon processor family. Intel software engineers work directly with Oracle Java engineers to identify and provide specific optimizations that take advantage of the latest Intel microarchitecture enhancements. Each new generation of Intel microarchitecture introduces enhancements that increase software performance. Older versions of Java will inherently run faster on newer platforms due to increased clock speeds and more efficient instruction execution. But to push performance even further, software engineers can also optimize Java to take advantage of new microarchitecture enhancements. When Intel releases a new microarchitecture and platform, Intel and Oracle software engineers work together to identify specific areas within the JVM that can be tuned to take advantage of the new hardware advances. These optimizations can deliver the highest performance possible across a broad range of applications, including Apache Hadoop.

Articles in this issue

view archives of Intel Software Adrenaline - Optimizing Java* and Apache Hadoop* for Intel® Architecture