The Profiling Service Module uses the multithreaded environment of the Data Integration Service. Therefore, the CPU speed is less important than the number of cores in the CPU. To calculate the number of cycles that the Profiling Service Module uses each second, add the clock speeds of the cores.

Profile operations run faster with more memory. When you run a profile on a flat file source, the Profiling Service Module uses memory to sort the value frequency data and buffer data. The Profiling Service Module performs multiple read operations to the same part of a file by reading from the memory buffer and not from the flat file on disk. This method applies to rule profiling for flat file and relational sources.

The Input/Output speeds for memory, disk, and network affect the performance of the Profiling Service Module. Higher speeds allow the Profiling Service Module to quickly access large amounts of data. Network speed affects the relational databases that are not on the Data Integration Service machine and flat files that are on a storage device attached to the network.