Lastly, you need to send the actual input data to the beamforming kernels and receive the output data from the beamforming kernels. This is accomplished through the PLIOs instantiated in the AI Engine application. Each PLIO specifies a 64-bit PL interface with an associated input or output data file. This means each row in the data files can store two (32-bit) complex data samples. An AI Engine kernel that is expecting a cint16 with a 64-bit PLIO needs a data file organized as four columns per row, where each column represents a 16-bit real or imaginary value.
<16-bit real> <16-bit imaginary> <16-bit real> <16-bit imaginary>