For embedded modules across backplane, I/O is serviced by the RPI rates set in the configuration. Minimum for each module is 1ms. However, it depends on number of modules and whether there are any specialty modules. As the number of modules increases, you have to increase the RPI or you may fault with overruns. Rockwells guideline is 2ms for 1 to 2 digital modules, 4ms for 3 to 4 digital, and 8ms for 5 to 6 digital. Analog and specialty add 20ms per module.