You can set up multiple threads and process one address at a time through each thread. This option offers the highest overall average processing speed by saturating all available FunctionServers specified by the NumFunctionServer parameter.

You can feed addresses in the engine with the bulk input feature. This feature allows simultaneous processing of multiple addresses (up to 100 addresses) through one or more threads. The engine distributes these addresses across all available FunctionServers. However, due to potential mismatches in address quantity or variations in processing time, some FunctionServers may remain idle while others process an address. As a result, the average processing speed reduces, especially when a large number of addresses (about a million) are processed.