Teradata Partners 2016 Recap
Teradata Partners held their annual conference this week in Atlanta. This is always a great event for me, full of learning and thought-generation as well as Teradata updates.
Teradata President and CEO (since June) Victor Lund was part of the gala kickoff and delivered a message essentially admitting the company was doing some reinvention. He classifies Teradata as an analytics solutions company with core offerings in business analytics, analytical architecture consulting and hybrid cloud products with a renewed focus on business solutions.
Teradata announced Teradata Everywhere, bringing its high-performance database to a suddenly large variety of platforms, especially in the cloud. Teradata is now available on Amazon Web Services, Azure (soon), Teradata Managed Cloud, VMware and the future of Teradata, Intelliflex.
Several items were presented in support of Teradata Everywhere. MAPS will provide customizable hash maps for data co-locating strategies and eliminating downtime during node expansion via table-level data distribution. The Teradata Database Adaptive Optimizer will now also adjust to the environment it resides in.
Under the Borderless Analytics initiative, Teradata QueryGrid and Teradata Unity were extended to Teradata Everywhere, as well as allowing any cross-platform queries should the need arise or make sense to be part of the architecture. Increasingly, users will be completely abstracted from the data platforms.
Teradata will be available in its massively parallel (MPP) form on AWS on up to 32 nodes and is available in most AWS regions now. Teradata Cloud also plugged a few holes in its availability, announcing Germany availability for Q32016.
The database itself is experiencing change as well. The 15.10 Database Adaptive Optimizer provides for in-stream query replanning. Prior restrictions on making tables columnar have been removed as well. Teradata Intelligent Memory has also improved with improved memory bandwidth and memory cache effectiveness. New load isolation levels and secure zones round out the database enhancements.
Aster also now runs on AWS as well as in Hadoop.
Every 10 years or so, Teradata announces a major change in its core platform. Ten years ago it was the Teradata 6400. That exciting moment is here again with Intelliflex.
Intelliflex (pictured) is on the upgrade path for all Teradata customers. It offers the ability to specify nodes at a detailed and heterogeneous manner so customers can, for example, separate storage and compute in node specification. Customers can also simulate any of the existing Teradata appliance configurations and change system characteristics over time. Intelliflex also offers double the performance density with far fewer nodes dedicated to stand-by, double the memory and all-SSD configurations.
Teradata also announced an expansion of investment in three “Connected Analytics” solutions: customer journey, financial foundation and early warning systems (IoT, predictive maintenance with sensor data) and new IoT accelerators.
Teradata is on the right track. The advancements are consistent with the resonating theme of Borderless Analytics. The customizable flexibility that will come with Intelliflex signals a bright upgrade path for Teradata customers and a compelling one for new customers.
Analytical ecosystems will get more complex before they become less complex. We are in a phase where data platforms are provisioned (quickly) for single workloads led by decoupled work units. Teradata Everywhere recognizes this and is bringing scale and performance in good old databases, combined with a strong accommodation for cluster approaches.
With Teradata Everywhere, Teradata is poised to deliver the data asset to the modern enterprise.