USPS Case Study
Analyzing big data with Intel® Xeon® processors helps streamline
operations, reduce costs, and drive new revenue opportunities
Enhancing Postal Efficiency
Delivering Real Time Analytics Case Study–SC13 Interview
Gerry Kolosvary, president of FedCentric Technologies describes at the SC13 Supercomputing Conference how the SGI UV delivered performance and scalability for his customer, United States Postal Service (USPS). The UV solution helped USPS move from historical analysis to real-time analytics and fraud detection. USPS faced an enormous challenge of processing over 4 billion scans daily, a challenge exceeds the one faced by the financial services industry.
Employing Memory Centric Database (MCDB) techniques, FedCentric’s High Density Computing (HDC) solution delivers up to 1.3M complex database updates per second and provided over 1000X performance increase over the incumbent disk based database solution. This breakthrough technology enabled USPS to identify and redirect fraudulent mail before it leaves the post office for delivery. The resulting savings paid for the system in the first few weeks of operation.
High Density Computing
Have you reached the limits of your commodity server?
High Density Computing combines commodity products,
software compatibility and user familiarity with “Orders of Magnitude”
Big Data / No Compromise
EXPERIENCE, LEADERSHIP, RESULTS
Our mission is to provide customers with the finest products and people to solve the most demanding "Big Data" applications
HIGH DENSITY COMPUTING
High Density Computing is an Intel Xeon (x86) and standard Linux (RHEL or SUSE) based system that enables in-memory processing of Oracle data sets from 64 GB to 48TB. This approach enables Oracle users to re-host applications from disk-based Oracle to MCDB with little or no changes to those applications.
"BIG DATA" SOLUTIONS FOR ORACLE 11-12 USERS
Memory Centric DataBase (MCDB) is an in-memory capability that enables Oracle 11-12 users to achieve order(s) of magnitude increases in "Big Data" ingest rates, query response times and real time analytics.