Briefs

Pervasive DataRush Knowledge Discovery Platform Redefines State-of-the-Art for Parallel Data Mining and Predictive Analytics

AUSTIN, Texas--(BUSINESS WIRE)--Advancing its emerging leadership in next-generation data-intensive applications processing and analytics, Pervasive Software(r) Inc. (NASDAQ:PVSW) today announced early availability of its revolutionary Pervasive DataRush Knowledge Discovery Platform, redefining state-of-the-art data mining and predictive analytics. Leveraging stunning results from initial projects, the Pervasive DataRush™ team is engaged with prospective customers to validate use cases and productize its high-performance data mining and predictive analytics capabilities.

"The Pervasive DataRush Knowledge Discovery Platform is based on Pervasive DataRush, launched earlier this year, and is designed to address the market's need for a massively parallel data mining platform that efficiently handles large, complex datasets, including those with thousands of variables and millions of records," said Mike Bryars, Pervasive DataRush General Manager. "Our groundbreaking Knowledge Discovery Platform incorporates leading-edge algorithms and runtime capabilities that enable highly scalable analytic applications and free developers from the constraints of dependence on memory-based computational models."

Although analytics is an emerging growth market, the vast majority of existing code is either single-threaded or parallel in very limited ways, making the entire data mining industry largely unprepared to fully exploit multicore capability and exploding volumes of data. The process of building analytic applications with existing industry technologies and large, complex data sets is very difficult and labor-intensive, resulting in slow development time and significant cost.

The Pervasive DataRush Knowledge Discovery Platform is multicore-ready with built-in features that dynamically parallelize every task through a combination of horizontal and vertical partitioning and pipelining - delivering unprecedented scaling on both data size and dimensionality, which in turn dramatically improves the ability to analyze large datasets. These powerful capabilities eliminate the constraints of processing limited samples of data, make parallel programming of analytic applications practical for the typical developer, and help expand the historically narrow usage of analytic applications from scientific and high-cost custom implementations to the mass market.

Read more.