Traditional systems lack the scalability and integrated architecture required to support big data analytics necessary in today’s business environment. Teradata and SAS® have joined forces delivering innovation to break through your big data analytics challenges and quickly uncover hidden opportunities for data-driven decisions. By integrating SAS with Teradata for in-database and in-memory processing of analytics, it is a fundamental paradigm shift to bringing analytics to where the data reside which helps to enhance performance, economics and governance.
Figure 1: Big Data Phenomenon
Teradata Continues to Lead the Pack
This is not the first time Teradata has driven a fundamental shift in data analytics. As early as 2010 Teradata was taking the data analytics world by storm with an industry changing shift in how data analytics was being delivered to organizations with the release of what was then called the Teradata Integrated Analytics Portfolio. This new Teradata analytics suite was used to deliver what they termed as an “Analytics for Everyone” message, built around a very tightly integrated and complete data analytics suite.
The family of capabilities inside their purpose-built platform covered the entire data management stack, including Teradata’s core database framework with a data mart appliance, a data warehouse appliance and a mix of Teradata embedded services, custom services and virtual machine capabilities. Teradata’s broad offering ranged from application development, native temporal data support, agile analytics acceleration, an advanced analytics suite optimized for in-database data mining, geospatial capabilities, OLAP optimization, unstructured analytics to visual data exploration capabilities across the entire suite.
In keeping with their trend of leading innovation in data analytics, in 2013 Teradata brought about a major shift in data analytics when they were one of the first major data management platform solution offerings with an integrated, scalable, fully parallel analytics platform built around the open source programming language, called R, and in-memory analytics libraries. Teradata was responding to their customers’ desire to leverage fuzzy logic (AKA fuzzy set theory) and the need to perform analytics on new data types such as XML, temporal and geospatial data.
Now we see another data analytics step forward taking place with the partnership between Teradata and SAS where the two market leaders have combined their strongest analytics and data management offerings, one they are calling “The New Movement: Bringing Analytics To The Data.”
An Important Change in Architecture Design
What do they mean when they say “bring analytics to the data” exactly? How does this fit into the brave new world of big data, and how does it benefit today’s organizations looking for ways to extract hidden value from their data?
The heart and soul of this new approach is built around a shift in where analytics workloads are performed. Traditional database and analytics workloads have been performed in the classic “client/server” model we’ve all come to know. In the traditional model, a “server” in the form of a central database environment is polled by slave systems to extract a subset of data through various means, such as structured query language (SQL) requests made by application software or software agents directly to the database engine, resulting in a stream of data being returned to and processed locally by the “client” system – reports or analytics would be produced by the local “client” system, and hey presto, you would have a result of some form on a screen or printed out on paper.
Unfortunately in this model, the power of the local client system where the analytics is being performed, more often than not a low-end desktop or laptop personal computer (PC) or workstation, is a significantly limiting factor due to limited processing capabilities, certainly not the type of high-powered processing required to perform quality analytics on the ever-increasing volume of information becoming available as organizations build big data capabilities.
The Teradata and SAS Partnership
To meet the growing market demand for big data era analytics support, Teradata and SAS have partnered to offer an end-to-end solution built on the best technology they have jointly. SAS® Analytic Suite for Teradata encompasses the analytical data lifecycle to help explore, prepare, develop and deploy the data model with a combination of in-database and in-memory capabilities.
Figure 2: SAS® Analytic Suite for Teradata
The SAS® Analytic Suite for Teradata leverages in-memory and in-database processing which has brought to the market a data analytics architecture designed around the logical shift to moving core data and analytics processing and decision calculations into the database engine. It is offering an entirely new approach to tackling the heady challenge of big data analytics in a cost effective and timely fashion for data-driven decisions.
In-Memory and In-Database Processing is a Game Changer
Customers globally are adopting the technology to improve performance, economics and governance. It is implemented in all industries that want to streamline their analytical process and enable organizations to make data-driven decisions more accurately and with better precision.
Teradata reports that the key customer benefits include:
- More accurate results leading to improved decision making
- Streamlined analytic data preparation steps to improve efficiency and reduce time to analytic results
- Faster model development and deployment to improve productivity and free up analytics team to tackle challenging issues
- Reduced data movement, redundancy and latency by leveraging in-database processing
- Meeting model governance and regulatory requirements by verifying scoring processes and auditability when called upon
In-memory processing is a new approach to tackle big data by using an in-memory analytics engine to deliver super-fast responses to complex analytical problems. The Teradata Appliance for SAS is a companion to the Teradata platform family executing in-memory analytics directly against the data, without having to move or duplicate the data, improving performance and data governance within the Teradata Unified Data Architecture (UDA) as shown in Figure 3.
Figure 3: Teradata Unified Data Architecture with SAS
The Teradata Appliance for SAS combines the strengths of SAS in-memory analytics and Teradata’s high-performance data management in a single, integrated platform. With this appliance, it can:
- Dramatically increase the yield from current analytic modeling processes
- Compress analytic modeling life cycles and visualization from weeks to minutes
- Solve business problems previously thought un-addressable in a timely manner
- Greatly increase the productivity of your analytic staff
Click here to view the briefing room webinar on this topic.