IBM Births New Server Type

IBM on Oct. 16 announced a new species of server: the IBM Information Server. The new server is all about grafting formally siloed information-handling products such as data cleansing and profiling platforms into one product with a common interface.

IBM unveiled the Information Server at its first Information on Demand Global Conference in Anaheim, Calif. The server packages up technology from IBM’s profuse acquisitions in this space, including Ascential, Venetica, Unicorn and CrossAccess.

It’s all about attaining a master set of information—what Oracle and now IBM have called a “single version of the truth,” said Tom Inman, VP, Marketing, IBM Information Management Software.

“One challenge with delivering information on demand is where this information sits,” Inman said.

“If you’re a retail bank, you want to know where your customer information sits, but … it can be scattered throughout the enterprise. … If you’re a government agency, you want to protect your citizens. It’s very difficult when information’s scattered through the enterprise like this.”

C-level executives have told IBM in surveys that they need to do a better job of leveraging their information, Inman said, yet they don’t believe they can do that given the current state of technology.

“Capabilities exist to do some of the important things, such as profiling and cleansing data, federating views, or transforming from one format to another, or being able to have metadata,” he said. The missing part is the ability to tie it all together to achieve the “one version of the truth,” he said.

IBM Information Server is designed to harness disparate data sources, to cleanse data, to transform and restructure information for its intended use, and to deliver data in the form needed.

In a press release, IBM gave an example of a likely customer: a manufacturer that needs to integrate data from subsidiaries, retail outlets and trading partners.

In such a case, Information Server could be used to combine sales figures with real-time inventory data from various sources, to standardize and merge customer data, and to deliver consistent information to operational data stores, data warehouses or master data applications.

Information Server will bring together disparate data-focused tools into a platform with one common interface and the ability to move data between the tools, all with a more scalable engine, said Judith Hurwitz, an analyst for Hurwitz & Associates. As such, it treats information as a service, as opposed to handling it as siloed products do.

“[IBM is] definitely starting a trend,” she said. “If you can take all the piece parts and work them together seamlessly, it really” would be a first, she said.

Competitors are on the same track. Oracle announced on Oct. 9 that it is buying Sunopsis, an ETL vendor whose technology will allow Oracle databases to pull in data from disparate data sources.

Click here to read more about Oracle’s Sunopsis buy.

Who will buy this new, high-end server? Steve O’Grady, an analyst with Redmonk, said that the past few years have seen an appropriate customer materialize.

“What we’ve seen in the database market in the past couple years is there’s a couple of different types of customers,” he said. “We have a customer who wants a simple database, one that’s easy to acquire and get up and running. That’s reflected in the traction MySQL has gotten in the market.

“At the other end of the spectrum are customers who’ve had a need for a platform to build from that allows you to leverage and take advantage of a number of different data types and sources. That’s the kind of need that a lot of the unstructured, or enterprise information integration, players, have targeted over the past couple years.

“IBM has determined … [there] is a market that says we have applications, being built on top of data, and we have data in lots of different places. What we need is an offering that solves that problem. An offering customers can build on top of. It’s a very, very different market than you’ll see with MySQL or what have you. But a market in its own right. The types and volume of data being generated keeps continuing to escalate. At the high end—and this is clearly a high-end product—is a need for a product that sits at the hub of all that. There’s a definite need.”

IBM Information Server has been through two beta tests with more than 75 IBM partners and customers. It is due Nov. 1. Pricing begins at $100,000 U.S. list price for an entry-level configuration.

Check out eWEEK.com’s for the latest database news, reviews and analysis.