Scalable Generic Logic Synthesis: One Approach to Rule Them All (Experiments for DAC'19)

Installation

• Install Python 3 binding (versionn 3.0a2.dev5) of cirkit3 (https://github.com/msoeken/cirkit) with pip:

python3 -m pip install cirkit==3.0a2.dev5

• Clone and build abc: https://github.com/berkeley-abc/abc (required for combinational equivalence checking)

• Add the path to the executable of abc to your PATH variable

• Clone the experiments repository (this repository)

• Run the experiments using Python 3:

python3 run.py 

Additional notes

Two parameters in the top section of run.py can be customized to show more/less information:

Parameter	Effect
verbose = True	Prints the statistics for each optimizing transformation
print_progress = True	Prints statistics for each benchmark (immediately when finished)

As a final results, the script run.py produces the data in Table 2 in [RienerTH+19] in LATEX format, i.e., the results of running compress2rs using AIGs, MIGs, XAGs, and additionally also for XMGs. The columns for XMGs are not shown in the paper.

Reference

These results are described in the paper [RienerTH+19]: Heinz Riener, Eleonora Testa, Winston Haaswijk, Alan Mishchenko, Luca Amaru, Giovanni De Micheli, Mathias Soeken, Scalable Generic Logic Synthesis: One Approach to Rule Them All, in Design Automation Conference 2019.