Making a battery of announcements at its annual Partners conference, Teradata on Monday reinforced its position as a leader in data warehouse with sweeping upgrades of its database, hardware and analytics strategy.
If 2009 marked the year in which Teradata filled out its hardware family, 2010 is mostly about filling out the company's vision for analytics. The strongest strategic message centered on Teradata Accelerated Analytics, which packages the company's diverse analytic capabilities and partnerships as a portfolio.
A few components of that portfolio are new, like temporal (time-based) insight and analyses of unstructured data sources, such as Hadoop and social media sources. Several components of the portfolio already exist but have been upgraded.
Support for OLAP analysis, for example, is not new, but a Terredata Aggregate Designer now recommends optimal indexes for Microsoft, MicroStrategy, Oracle, SAP BusinessObjects and SAS OLAP environments. In addition, a new Excel plug-in gives Excel Pivot Tables direct access to Teradata, giving end users self-service access to data.
Underscoring the company's independence, the Teradata Accelerated Analytics portfolio reaches out to an extensive community of partners and third-party vendors. For example, Teradata supports a growing set of in-database processing approaches for SAS, IBM SPSS and KXEN commercial analytics. This latest release specifically steps up support for open source R tools and analytic models.
At the center of the news is Teradata 13.10, the latest release of the vendor's database, which supports improved compression and temporal (point-in-time) analyses. Improved compression is a frequent and always-welcome improvement in databases as it lets organizations cram more data into existing deployments. New block-level and algorithmic compression are said to work together to increase routine compression up from 2X to 3X into the 3X to 4X range. That matches rates recently claimed by competitors EMC Greenplum and Netezza.
The temporal advances in 13.10 make it a time-aware database that can track changes over the history of data.
- Insurance companies need to understand who had policies when and when they changed coverage levels; otherwise, they can't bill customers and handle claims appropriately.
- Manufacturers need time-based analysis to gauge sales performance. How do you fairly calculate commissions or bonuses when territories have changed in the middle of a month or year?
- Retailers have to understand brand and category performance amid change. If ice cream is moved from frozen foods to dairy, how do you accurately measure year-end performance of each department or related suppliers?