In-Depth

At Partners Conference, Teradata Reloads

Teradata likes to use its Partners conference as a launching pad for its splashiest product announcements. This year's event was no exception.

• By Stephen Swoyer
• 11/03/2010

Teradata Corp.'s Partners user conference always produces a rush of data warehousing-oriented news items. Not only is Teradata one of the biggest names in the enterprise DW space, but Partners tends to attract lots of other big names.

Many of Teradata's partners are on hand, for one thing: vendors such as MicroStrategy Corp., SAS Institute Inc., Informatica Corp., Information Builders Inc. (IBI), and Kalido. Most of Teradata's nominal competitors -- e.g., IBM Corp., Microsoft Corp., Oracle Corp., and SAP AG -- are exhibiting, too.

Teradata likes to use Partners as a launching pad for its splashiest product announcements. For example, Teradata unveiled a new version of its flagship database (Teradata Database v13.10) -- complete with a new "time-aware" or "temporal capability;" shuffled (and revamped) its DW platform family, showcased a new Unified Logical Data Model (LDM) framework; sought to reposition its warehouse systems as "Analytic Services Environments;" and announced a new version of its Teradata Marketing Management Suite.

That wasn't all. At previous Partners events, several Teradata Partners trumpeted product announcements of their own.

Teradata Reloads

Teradata officials like to claim that Teradata Database is at the top of the DW industry's evolutionary pecking order. With a few exceptions -- and more than a few qualifiers -- Teradata's competitors generally concede this point. Several obvious exceptions include Hewlett-Packard Co., which positions its own Neoview platform as roughly analogous to Teradata's line. IBM, one of Teradata's most stalwart competitors, recently upped the competitive ante with its acquisition of Netezza Inc. Oracletouts its Exadata systems as Teradata and Netezza killers.

The v13.10 release of Teradata Database ships with a new "time-aware" facility that officials claim can make it easier (or faster) to track changes in data over time. It's the difference between an incomplete, point-in-time picture -- in which an analyst assesses information in the context of a delimited "slice" of time -- and the much-ballyhooed "complete picture" that DW players always like to trumpet. Teradata's new temporal facility builds on the "PERIOD" type that it first introduced with v13 of Database last year. PERIOD permits Teradata users to store temporal data in a single column and at the same time tap operators and functions to manipulate it.

Other v13.10 improvements include new compression options and new analytic functions in Teradata's Open Parallel Framework.

On the platform front, Teradata announced several new DW systems based on the latest silicon from Intel Corp. The first such entry, Teradata's Active Enterprise Data Warehouse 5650 system, is said to boost performance by up to 43 percent. (This system supplants the Active Enterprise Data Warehouse 5550 system at the top of Teradata's platform hierarchy.)

The new Teradata Extreme Performance Appliance 4600 likewise ups the ante in the appliance space, where -- since it first introduced its Extreme Data 1550 two years ago -- Teradata has been able to compete at the $16,000-per-TB price point. The Extreme Performance Appliance 4600 is slotted for a higher-end niche, however. Teradata bills it as an industry-first: a DW system based entirely on solid state disk (SSD) storage.

Other new appliance offerings include the Teradata Data Warehouse Appliance 2650 (an improved version of Teradata's existing 2550 system that ships with an available "Analyst Pack"); the Extreme Data Appliance 1650 (an appliance that supplants its 1550-branded predecessor) and the Data Mart Appliance 550 (another beefed-up appliance that likewise supplants its 550-branded predecessor).

Re-Birth of a Data Model?

Teradata says its new Unified LDM was honed at hundreds of customer sites and is the product of a decade of product development. It's designed to make it easier for customers to model and integrate processes or data from internal or external sources. The concept builds on an established best practice in Teradata environments -- that of customizable data models specific to certain industries. Teradata promises that Unified LDM can give customers the best of both worlds: the benefits of an industry-specific data model rooted in common foundational and functional modules.

"Our customers operate in value-creating networks that cut across standard industry frameworks. They increasingly acquire and share information across their supply chains, demand chains, and channel Partners, as well as with other stakeholders," said Tom Russell, vice president of architecture and modeling solutions with Teradata, in a statement.

"[T]here is some value in exchanging information through isolated point solutions," Russell continues, noting that "there is far greater value in integrating all of the information into a data warehouse supported by a cross-industry data model." Teradata Unified LDM, he argues, enables customers ""to see beyond the enterprise view to a holistic 'value network' view of their business."

Repositioning

In an intriguing bit of repositioning, Teradata used Partners to recast its data warehouse systems as "Analytic Services Environments," touting its platform-wide support for in-database analytics -- along with its support for new data types and emerging data paradigms; third-party facilities; and application development interfaces. In the latter case, for example, Teradata offers plug-ins for two popular integrated development environments (IDEs): Microsoft's Visual Studio .NET and the open source software (OSS) Eclipse IDE. These plug-ins permit developers to expose MapReduce-driven analytics via drag-and-drop interfaces.

Teradata has long provided support for geospatial, OLAP, and other kinds of in-database analytics. More recently, it partnered with several vendors (including both SAS and KXEN) to promote advanced analytics -- e.g., the ability to run SAS or KXEN applications on top of the Teradata database.

"The challenge with today's analytic environments is that data and analytics are often siloed across many diverse systems. Analysts must often move large amounts of data between various environments to be able to execute the complex analytics required to meet business demands," said Teradata chief development officer Scott Gnau, in a release. "With Teradata Accelerated Analytics, customers are able to integrate analytics and data, dramatically reducing time-consuming data movement while executing analytic parallel processing in-database for superior performance."