MemSE

This repository contains a Pytorch implementation of MemSE as discussed in the paper:

• "MemSE: Fast MSE Prediction for Noisy Memristor-Based DNN Accelerators" published in AICAS2022

If this project is useful for you, please cite our work:

@inproceedings{kern2022memse,
 title={MemSE: Fast MSE Prediction for Noisy Memristor-Based DNN Accelerators},
 author={Kern, Jonathan and Henwood, Sebastien and Gonçalo, Mordido and al.},
 booktitle={Proceedings of the IEEE International Conference on Artificial Intelligence Circuits and Systems (AICAS)},
 pages={TBA},
 year={2022}
}

Usage

1. Environment set-up and dependencies

• Python 3.8
• Libraries (Pytorch, Numpy, SciPy, Tensorly and opt_einsum)

To install from source, run the following commands:

git clone https://github.com/sebastienwood/MemSE.git
cd MemSE
python setup.py install

2. Paper experiments

All the papers experiments can be found under experiments/aicas.

License

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

The software is for educational and academic research purpose only.