Inside IBM's Push Into Real Time Predictive Analytics
I attended an event for analysts and media on Tuesday at IBM's Thomas J. Watson Research Center in Hawthorne, N.Y., where the company launched its Smart Analytics System.
However the news of its new offering and future roadmap was drowned out when IBM announced that it has agreed to acquire SPSS, a leading provider of real time predictive analytics software, for $1.2 billion (see original story here). Pending shareholder and regulatory approval, the deal is slated to close by year's end.
Chicago based SPSS is regarded as the leading provider of real-time predictive analytics software. With 1,200 employees, its technology is widely used by enterprises worldwide. During the event, I sat down with Forrester analyst James Kobielus, who said this is a significant move by IBM. "The SPSS acquisition is strategic for IBM," Kobielus said.
While IBM last year gained a leading business intelligence and analytics portfolio via its $4.9 billion acquisition of Cognos, the key component IBM lacked in its Information on Demand (IOD) offering was a best of breed data minding offering. "This fills out the portfolio," Kobielus said. In a blog posting, he described the deal as a "bold move has already sent shockwaves throughout the analytics market." You can read his entire reaction to the SPSS deal here.
SPSS is the second largest provider of predictive analytics data mining statistical analytics tools, he noted. The largest is SAS Institute. He pointed to few overlaps with SPSS such as IBM's DB2 Intelligent Miner within the InfoSphere portfolio, though he predicts that will be phased out as IBM builds out the SPSS brand within its IOD portfolio.
"Fundamentally their technology is componentized so we can embed it anywhere," said Amuj Goyal, general manager of IBM Software’s information management software organization.
It remains to be seen whether IBM will continue SPSS integration with other data warehouse providers including Oracle, Microsoft, Sybase, and Teradata. "I doubt IBM will rock the boat," Kobielus told me, saying its in its interest to keep SPSS offerings heterogeneous.
Meanwhile, the news that got drowned out by the deal was the launch of the IBM Smart Analytics System (IAS).
The system consists of an IBM pSeries server based running AIX, that includes storage, networking, various other services and a suite of IBM’s data mining tools including its DB2 database, Cognos BI and InfoSphere Warehouse. It will be available in September with a starting configuration of 4 terabytes and up to 200 Tbytes. Pricing was not disclosed.
Kobielus said IAS extends IBM's existing portfolio of data warehousing appliances. "What it adds is pre-integrated business content geared to particular vertical and horizontal markets. So it includes DB2, the data warehouse appliance, IBM Information Server, data integration and design tools plus the application specific or vertical specific dashboards and workflows and meta data and cleansing tools."
What does IAS mean if you're a developer? "IBM is very much turning its data warehouse portfolio into an application server in that they are pre bundling all of these solution components and providing application development interfaces to allow ISVs to build targeted applications on top of IAS," Kobielus said. "It's really a development platform, and ties into a services oriented architecture. IBM hasn't really called out that theme but it's undoubtedly part of their road map."
IBM said it will release by year's end technology it calls an Analytics Optimizer, which combines hardware and software to perform even faster analytic queries. "We are doing in-memory exploitation, we are exploiting vector processing inside this predictive optimizer, we are evaluating predicates in parallel using new scanning technologies," Goyal explained.
The goal, said Steve Mills, senior vice president and group executive for IBM Software Group, is much faster queries with the target of real time business automation "One hour queries for many people don’t cut it," Mills said. "Five second queries all of a sudden start to open up the aperture to more creative thinking."
Posted by Jeffrey Schwartz on July 30, 2009