Influencing Factors

The SAPS number as shown by SAP benchmarks can be influenced by a multitude of factors. Indeed, it is not a pure hardware performance indicator, but also a software performance indicator—measuring a Unicode system yields lower SAPS numbers than a non-Unicode system. On the hardware side, it is not only about the processor. Increasing the count of memory DIMMs built into a server can decrease its SAPS value because memory access frequency may go down. Decreasing the amount of memory in a server may decrease its SAPS number because of smaller buffer sizes, which is the reason why using large DIMM sizes to populate a SAP server's memory banks is preferred.

Benchmarking a Unicode SAP system can deliver a lower SAPS number than a non-Unicode system (as long as all other factors stay the same). Because most new SAP installations are Unicode-based, Unicode has been mandatory for SAP SD Standard Application Benchmarks since the beginning of 2009. This requirement means it is important to pay attention to SAPS numbers from 2008 in comparison to SAPS numbers from 2009.

Even the year of a benchmark run indicates how to interpret a SAPS number. Starting in 2009, SAP SD Standard Application Benchmarks must have a response time below one second (instead of 2 seconds before), must be Unicode-based, must use the general ledger business functionality, must use Enhancement Package 4 for SAP ERP 6.0, and must use special versions of SAP_APPL and SAP_BASIS. Because of these changes, the same hardware can deliver a lower SAPS number in benchmark certificates of 2009 than of 2008.

In the end, the SAPS number heavily depends on the person actually doing the benchmark. Is the person trying to get the highest number possible, or does he or she try to get real-world figures so that a customer can just take them and be on the safe side regarding his hardware's capabilities?

Bottom Line

While the SAPS standard is a great way of mapping your business needs to available hardware, there are some drawbacks to using it for comparing hardware. If two servers deliver different SAPS values, the reason can be in different benchmark rulesets, different strategies (real world versus max out), different memory configuration, and different SAP configurations during a benchmark. As long as you do not have an exact knowledge about this, one SAPS value will still be very fruitful for you and that is your system's requirement in SAPS. Use it as a discussion base when talking with the Dell SAP Sizing Experts.

See also

http://sap.com/benchmark