The Boulder BI Brain Trust


Richard Hackathorn: October 2009 Archives

IBM Launches Smart Analytics for the Smart Planet

| | TrackBacks (0)
IBM logo.pngIBM Corp discussed their directions in Data Warehousing with Bill O'Connell, DE, Chief Architect IM Solutions, Warehousing CTO and Greg Lotko, IBM Software Group, Information Management, Vice President, Warehouse Solutions.

sas logo.pngGreg started by giving an overview of what clients are telling us... single analytics system with faster time to value and reduced cost of ownership, and greater access to information. IBM announced IBM Smart Analytics System in July with GA in a full-stack "analytics" system. It is like an "appliance" with one product number, one delivered package, one support point, 3 years of maintenance, initial training, etc. It could be called "Cognos in a Box" leveraging IBM's acquisition of Cognos. By packaging these capabilities, there is a 2x pricing advantage as compared to buying the separate hardware/software/service/support components. The system can increased capacity along several dimensions. Storage starts at 4 TB and can grow in 4 TB increments. For the smaller configurations, IBM has inventory and can deliver in two weeks.

As a different product, Smart Analytics Optimizer is a backend X-series hardware rack to optimize analytic workload for System z. The product was announced in July. Greg indicated that "first customer ships by end of year". As a back-end enhancement, the  customer infrastructure is unchanged. An interesting aspect is that the data persists on the optimizer so that each analytic request munches on the same resident data.

Bill discussing the value chain of IBM Information Agenda. Reviewing innovations over last 40 years! Then discussed the eXtreme Analytics Platform (XAP) that takes wide spectrum of unstructured data from diverse external sources into a MapReduce HADOOP data engine. IBM's motivation is to provide corporate IT with a value-enhanced alternative to each user doing their own thing on their own personal data cache from unstructured external sources. IBM has proposed JAQL as an open language for analytics processing on XAP.

Ibm arch2.pngBill continued by explaining the modular architecture behind the Smart Analytics System, showing expansion in data, user and fallover modules, along with add-on modules for Cognos and InfoSphere Warehouse modules.

Bill argues that the future technologies that will enhance the Smart Analytics System will be disruptive to the marketplace. For example, solid-state disk racks will result in 5x in performance, 2x fewer disks, and 2x better energy consumption. Bill mentioned the trend in data mining shifting from specialized projects by statisticians to application servers by application developers.

My Take
... IBM has positioned itself as a major vendor for analytics on enterprise data warehouse with an impressive array of capabilities. Those of us that have mucked around the DW/BI industry for decades often lose perspective on the extent that we have matured. This is a good example of industry maturity. The focus has shifted from how many TBs are stored to how complex analytics are generated for user-facing applications. This is progress toward DW/BI technology driving real business value. 


This page is a archive of recent entries written by Richard Hackathorn in October 2009.

Richard Hackathorn: August 2009 is the previous archive.

Richard Hackathorn: November 2009 is the next archive.

Find recent content on the main index or look in the archives to find all content.