At the IBM Impact 2014 conference this week, IBM provided a glimpse into two strategy initiatives that ultimately promise to change how data is consumed and managed across the enterprise.
Looking to give end users more direct control over data, IBM previewed a Blue Insight offering through which end users will be able to leverage the forthcoming BlueMix cloud integration service to integrate data without the aid of a developer. The Blue Insight offering consists of a data refinery that acts as a virtual data repository, and a set of Catalyst Insight tools will enable end users to analyze any source of data connected to BlueMix.
At the same time, IBM is planning to significantly optimize in-memory performance of its DB2 database by finding a better way for relational stores for transaction processing applications and columnar stores for analytics applications to coexist inside DB2. Sean Poulley, vice president of information management for IBM, says a version of BLU Acceleration that IBM will have ready this summer will smoke the pants off rival in-memory databases such as the SAP HANA.
In addition, Poulley says that IBM will be adding data virtualization capabilities that will make it easier for organizations to process data where it resides within the context of a larger virtual data warehouse.
Taken together, these forthcoming initiatives are designed to provide access to data at an unprecedented level of scale while taking advantage of in-memory and magnetic storage systems to maximize performance. Poulley says that while memory is becoming an important part of the data warehouse, it doesn’t make economic sense to store all data in Flash. The real challenge, says Poulley, is to allow relational and columnar data stores to take advantage of existing and emerging memory technologies in a way that allows them to scale now and well into the future.
IBM, SAP, Oracle, Teradata and others are locked in a bitter fight to gain the inside edge when it comes to driving the next generation of data warehousing technologies in the enterprise. For the most part, those data warehouses will be defined by Hadoop, serving primarily as a data hub to in-memory platforms that can handle the most complex analytics queries in seconds.
While it’s not clear which vendors will ultimately win that contest, the one thing that is for certain is that the data warehouse as we once knew it will never be the same.