Actual development is being done in KastnerRG/hls4ml in the branch aie_backend/dev. That branch is included here as git submodule
git pull --recurse-submodules conda env create -f environment.yml conda activate hls4ml-tutorial python quickstart.py cd golden python run_workload.py source /tools/Xilinx/Vivado/2025.2/Vitis/settings64.sh cd aie_pl # edit gen.py to set layers make sim # AIE : 570 cycles (456.0 ns @ 1.25 GHz) [aiesimulator] # PL : 127 cycles (406.4 ns @ 312.5 MHz) [HLS report] # Total (analytical): 862.4 ns → 9.28 M samples/s make run_emu