IBM will step up its "big data" game Wednesday when it unveils a new edition of its PureSystems converged infrastructure servers for Hadoop implementations and big data-related enhancements to its DB2 flagship database.
The move comes shortly after IBM projected that its annual sales of data analytics products would reach $20 billion by 2015, up from an earlier estimate of just $16 billion.
The new offerings are the latest steps in IBM's efforts to develop a deep portfolio of big data products, according to Bob Picciano, general manager of the company's information management operations.
"In the last five years, IBM has invested billions in research to unlock the power of big data, and we're delivering some of the resulting innovations today," he said in a statement to be delivered at the IBM Research-Almaden facility in San Jose. "Increasingly, the demand for these capabilities is moving from the back office of large enterprises to public-facing engagements with customers, citizens, buyers, partners and employees, and that means data needs to move at the speed of business."
The new IBM PureData System for Hadoop is an addition to the PureSystems line that's designed to make it easier for businesses to implement and manage Hadoop-based systems. Hadoop is an open-source platform from the Apache Software Foundation that's used to collect and manage huge volumes of structured and unstructured data.
IBM introduced the PureSystems line one year ago and the company has sold 2,300 units. The line includes the PureApplication, PureFlex and PureData systems.
The new PureData System for Hadoop is an extension of IBM's InfoSphere BigInsights, the company's pre-packaged Hadoop offering, said Nancy Kopp-Hensley, director of IBM's big data product strategy, in an interview with CRN. The system has a built-in data archive capability that makes it possible to analyze historical data along with real-time data analysis.
The PureData Systems for Hadoop is expected to begin shipping in the third quarter.
IBM is also set to debut several data management and analysis technologies under the name "BLU Acceleration." The technologies provide ways to parallelly analyze data across multiple processors, and identify and skip over data that's not relevant to a query, said Bernie Spang, strategy director for the IBM Software Group, in an interview with CRN.
NEXT: Performance Increases, New Software ReleasesA new "actionable compression" feature lets users analyze compressed data without decompressing it. And, new in-memory processing technology makes it possible to leverage in-memory compute systems even when the amount of data exceeds memory size.
The new technologies altogether will provide performance increases of 8x to 25x for reporting and business analytics applications, Spang said. They are being built into upcoming releases of IBM's Informix and DB2 databases, followed by other IBM software products.
IBM also will unveil new releases of the company's InfoSphere Streams software for analyzing data streams in real time, Informix Time Series software for reporting and analyzing time-stamped data, and the InfoSphere BigInsights distribution of Hadoop.
Kopp-Hensley said the new products and product enhancements should appeal to IBM's channel partners that focus on data management and business analytics.
PUBLISHED APRIL 3, 2013