SAS^® In-Database Processing

Customer requirements to understand and act on critical business issues such as fraud detection, credit risk, price optimization, warranty analysis and customer retention demand efficient handling and utilization of business intelligence (BI), data integration and analytic components. The scope of analytic computations and the volume of data and data sources are growing at an unprecedented pace. Enterprises need flexibility to manage the analytical life cycle, from discovery to the execution of large numbers of new and existing analytic models that are designed to address functional and industry-specific business issues in a secure, scalable manner.

SAS has always shown that analytic workloads are distinctly different from transactional processing. We have been working with relational database management system (RDBMS) vendors for years to help mature their ability to handle mixed workloads. Relational database vendors have slowly gained healthy adoption at the data warehouse layer as their products begin to handle mixed workloads (i.e., analytic queries/workflows and transactional queries), high availability and right-time "refresh" rates.

The goal of this initiative is to offer our customers innovative solutions, delivering better answers faster. SAS will continue to support customers who want to keep up with the development, management and processing of analytic functions/workflows outside of the RDBMS. Customers who need to further leverage their investments in mixed workload RDBMS platforms will benefit from our road map to integrate and run SAS analytic processes within the relational database.

Currently, the best practice for analytic solutions is to build analytic/subject-focused data warehouses. For performance and flexibility, this is often the best choice. Sometimes, however, this creates challenges such as:

• Time constraints in moving large volumes of data.
• Management and provisioning of data.
• Proliferation and governance of data.
  

Our support for In-Database processing will address these challenges by moving the relevant analytic tasks closer to the data and improving the integration between the SAS and database management systems (DBMS). The ability to perform these tasks from inside the DBMS will significantly reduce bottlenecks that result from moving data over a network. Keeping in mind the respective strengths of SAS and relational database technologies, the key benefits customers stand to gain are reduced data movement, improved performance run times, and faster analytic development and deployment to turn information into usable insights.