Unisys ES7000 servers, SAS report superior performance gains in ETL benchmarking

Enterprise Networks & Servers, May 2005

SAS, a provider of business intelligence and analytics, has announced superior performance of Unisys ES7000 servers running SAS software, citing the first hardware platform benchmark for SAS Enterprise ETL Server.

ETL (Extract, Transform and Load) is the foundation of business intelligence (BI), performing the resource-intensive "heavy lifting" for analytics and BI applications, which are only as effective as the information available in data warehouses and data marts.

For example, large financial institutions must access diverse data stores, update and manage growing amounts of data, perform updates more frequently with a smaller window of time, and cleanse and transform data into consistent intelligence.

As all this data grows, the window of time to process it does not. Therefore, organizations need enough power to process larger amounts of data within very limited amounts of time.

This benchmark is the first of its kind for SAS and will begin a series of benchmarks on all SAS hardware platforms.

Through collaborative benchmark testing and engineering, Unisys and SAS lower implementation risk by validating optimal performance, as well as by creating best practices in system sizing and configuration to meet SAS' high-performance requirements. These test results prove that SAS Enterprise ETL Server running on an ES7000 server can address very large-scale ETL processing requirements on a single Microsoft Windows platform.

"Unlike transactional processing environments, which are predictable and measurable by nature, business intelligence requirements are highly variable and unique to each organization," said Don Hatcher, vice president of the SAS Enterprise Excellence Center. "In evaluating and choosing the right enterprise intelligence platforms, it's essential to examine and optimize all aspects of the process for initial and long-term performance consistency. That's where these benchmarking results prove the most valuable."

Delivering the power of SAS across an enterprise means addressing a range of business issues and technical demands. Today's ETL processes need to support global business needs in a 24-by-7 operation.

Therefore, updates need to be completed in much smaller timeframes. Decision makers responding to competitive pressures and customer demands need information in near-real time. ETL performance is a critical success factor for enterprise business intelligence.

Complete benchmark results are available at unisys.com/es7/etl-benchmark.