Deep learning-driven EDA
Deep learning-driven digital circuit recognition
Publication
- Arash Fayyazi, Soheil Shababi, Pierluigi Nuzzo, Shahin Nazarian, and Massoud Pedram. Deep Learning-Based Circuit Recognition Using Sparse Mapping and Level-Dependent Decaying Sum Circuit Representations. In IEEE DATE, 2019.
Developers
Arash Fayyazi fayyazi@usc.edu
Pierluigi Nuzzo Nuzzo@usc.edu
Shahin Nazarian shahin.nazarian@usc.edu
Massoud Pedram pedram@usc.edu
Questions or Bugs?
You may send email to fayyazi@usc.edu for any questions you may have or bugs that you find.
%%%%%%%%%%%%%% ABC %%%%%%%%%%%%%%%%%%%% To compile ABC as a binary, download and unzip the code, then type
make for more information, visit https://github.com/berkeley-abc/abc
%%%%%%%%%%%%%% framework in ABC %%%%%%%%%%%%%%%%%%%% To compile the proposed framework: first go under abc/src/, create a folder called "ext"
under abc/src/ext, copy benchmarks and Sports directory.
under abc/, execute
make -j
Then in ABC folder, execute
./abc this command runs the binary in the command-line mode: then execute,
&r mul_4x4_a.aig; &if -K 4; sportCNN -F mul_4x4_a -M 1
if you want to generate CNN input data from all benchmarks, use following commands:
python ReadDataArash_baseline.py
or
python ReadDataArash_2L.py
or
python ReadDataArash_CellLib.py
based on desire approach. %%%%%%%%%%%%%% CNN run and test %%%%%%%%%%%%%%%%%%%%
Please first correct all paths based on where you put ABC and these files.
for training run following command:
python ArashCNN.py
or
python ArashCNN_2L.py
or
python ArashCNN_CellLibTest.py
Then for testing a input:
python TestRun.py input_file label
example:
python TestRun.py div_16d16_a 0
result will be:
CNN predict: division real operation is: division