A new white paper titled "Evaluating and Comparing Oracle Database Appliance Performance" was recently released. This white paper outlines a quick and simple process to evaluate the performance of Oracle Database Appliance for both OLTP and DSS workloads and compare it to the same workloads when run in your local environment. The testing for this paper was conducted using Swingbench's Order Entry (OLTP) and Sales History (DSS) workloads. Users can review already the documented results and conduct a reverse-POC (proof-of-concept) in their legacy environment by running Swingbench and then simply compare relative performance.
For the testing conducted during the course of writing this white paper, Oracle Database Appliance was able to support up to 10,000 Swingbench users for the OLTP workload (while maintaining sub-second response times) and provided throughput of more than 2400 MB/Second for the DSS workload. Both are significant numbers. The white paper provides the results of various core configurations tests (related to the pay-as-you-grow feature) in a matrix form.
The paper also covers some minor tweaks to the standard configuration that can enable even greater workload performance. I invite you to take a look at http://www.oracle.com/technetwork/server-storage/engineered-systems/database-appliance/documentation/oda-eval-comparing-performance-1895230.pdf