Posted by Chris Villinger, VP Business Development & Marketing
ScaleOut Software joins forces with IBM to help POWER8 system users achieve unprecedented speeds for live, fast-changing, big data analysis in Hadoop environments.
In the following IBM Systems ISV testimonial video, ScaleOut Software CEO and Founder, Dr. William Bain, describes the benefits of optimizing ScaleOut Software’s in-memory computing technology on the IBM POWER8 platform: POWER8 and ScaleOut Software: In-memory computing for operational intelligence.
The alliance with IBM is a marriage between two best-in-class, high-performance hardware and software technologies optimized for low latency and real-time execution. The POWER8 system offers unique capabilities in its hardware and a highly optimized Java execution environment. ScaleOut Software’s in-memory computing technology brings techniques from parallel supercomputing and big data to live applications, and we have optimized our software to take advantage of the POWER8’s breakthrough architecture.
The combination of ScaleOut hServer®, an in-memory MapReduce execution platform and IBM POWER8 equips organizations with the ability to track and analyze constantly changing data, empowering them to make more informed decisions in the moment. For example, benchmark testing with a financial services application showed that ScaleOut hServer, on an IBM POWER8 cluster, can track 20M positions, analyze strategies, and alert traders in under 3.4 seconds, in contrast to the 15 minutes or more that conventional techniques require. ScaleOut hServer can track real-time changes to investment strategies throughout the trading day and intelligently identify imbalances on a portfolio-by-portfolio basis.
Unlike other in-memory computing and data-parallel execution engines, ScaleOut hServer provides three distinct advantages. First, ScaleOut hServer’s in-memory data grid holds live, fast-changing data that can be simultaneously updated and analyzed at in-memory speeds. Second, ScaleOut hServer reduces the overhead associated with standard Hadoop distributions, by avoiding unnecessary batch scheduling, disk access, data shuffling, and mandatory key sorting. Finally, ScaleOut hServer on IBM POWER8 allows IT departments to run MapReduce jobs with zero code changes, streamlining development and accelerating time to results.
This powerful combination represents the future of in-memory computing, giving customers the ability to analyze live, fast-changing data as it fluctuates in real time. With ScaleOut hServer running on the POWER8 system, customers now can run in-memory MapReduce applications faster than ever and for the first time also on live, fast-changing data.
* IBM and POWER8 are registered trademarks of IBM Corporation.