The software design in the FIR Filter AI Engine implementation consists of the following sections:
AI Engine Kernels and Graph Representation
AI Engine Kernels and Graph Representation
DSPLib FIR filter kernels are C/C++ programs written using specialized intrinsic calls that target the VLIW vector processor. The AI Engine compiler compiles the kernel code to produce an executable ELF file for each of the AI Engines used in the design. Review AI Engine Kernel Programming Section in the AI Engine Documentation for a high-level overview of kernel programming. You can stitch DSPLib kernels together to function as AI Engine graphs written in C++. In this design, the AI Engine compiler writes a summary of compilation results to build/fir_aie_$(N_FIR_FILTERS)firs_$(N_FIR_TAPS)taps/Work/fir_aie_graph.aiecompile_summary. You can view the graph by running the following command:
vitis_analyzer build/fir_aie_$(N_FIR_FILTERS)firs_$(N_FIR_TAPS)taps/Work/fir_aie_graph.aiecompile_summary
The following figures show the graph representation of the AI Engine kernels (N_FIR_FILTERS=1, N_FIR_TAPS=64).