New SAS High-Performance Analytics allow customers to model entire data sets in-memory - not just a sample - on a dedicated Teradata appliance. By dramatically improving the speed and accuracy of analytical modelling for statisticians on big data issues, SAS High-Performance Analytics on Teradata allow customers to get a complete view of their business for improved decision making.
"Leveraging in-memory processing with the power of the Teradata architecture to minimize traditional I/O constraints, SAS and Teradata can now provide answers on big data problems across industries at the lowest levels of granularity. This leap in technology innovation will allow companies to do such things as more accurately monitor fraud, predict failures and provide real-time customer insights and offerings in ways they previously couldn't", stated Keith Collins, SAS Senior Vice President and Chief Technology Officer.
For example, in the financial services industry, in order to predict the probability of loan defaults, an analyst would build a predictive model from an extremely large set of data. Because of the large data size and movement constraints, the analyst most likely would use a sampling of the data to build the loan default model. This impedes speed of building the model and model accuracy, which means the bank's profitability on its loan portfolio is not optimized. By implementing SAS High-Performance Analytics on a Teradata appliance, these data constraints are eliminated, model development time is reduced from days to minutes, and loan default accuracy can increase by 10 percent to 50 percent, significantly improving the bank's profitability for its loan portfolio.
"The demand for operational analytics against integrated data from across the enterprise has become table stakes for leading businesses to achieve and maintain a sustainable competitive advantage", stated Scott Gnau, Chief Development Officer at Teradata. "SAS High-Performance Analytics helps our customers meet this demand by leveraging Teradata parallel processing to minimize the duplication, complexity and movement of data while increasing the performance of the analytic teams and impact of the business insight they generate."
James Taylor, CEO of Decision Management Solutions, faculty member of the International Institute for Advanced Analytics, and author, added: "Analytic value is lost when limited computing resources mean that data is excluded or that the time to get results is extended. When a business can ramp up its processing power and analyze all the data needed for a decision, it can maximize the value of analytics."
The new SAS High-Performance Analytics is part of the SAS High-Performance Computing portfolio, which spans across in-database analytics, grid computing and in-memory analytics, and will be generally available on the Teradata appliance in fourth quarter 2011.
This new offering is the most recent addition to the growing portfolio of analytic solutions from SAS and Teradata. Many features and functions of this portfolio of products have been influenced by the SAS and Teradata joint Product Advisory Council (PAC), which provides customer input and validation to joint offerings before they go to market. This influence has resulted in nearly 400 global customer engagements since the partnership started and has allowed the partnership to enhance and grow its suite of Advantage Programs to solve customer problems. In addition, the SAS and Teradata Business Analytic Innovation Center (BAIC) has also been integral in helping customers identify analytic best practices with domain and subject-matter experts, as well as quickly uncovering innovative models that provide unique insights for optimizing business operations.
The announcement was made at SAS Global Forum, the world's largest gathering of SAS users, attended by more than 3,500 business and IT users of SAS software and solutions.