Nerual Network of Stochastic Computing for MNIST Recognition

A systolic array simulator for multi-cycle MACs and varying-byte words, with the paper accepted to HPCA 2022.

This is a general-purpose simulator for unary computing based on PyTorch, with the paper accepted to ISCA 2020 and awarded IEEE Micro Top Pick for 2020.

Stochastic Computing for Deep Neural Networks

Fully Hardware-Based Stochastic Neural Network

An example of SC-FCNN on MNIST. This work is an un-offical implementation of paper: Dynamic Energy-Accuracy Trade-off Using Stochastic Computing in Deep Neural Networks. We use full-SC process, all values are representated as bit-stream. and got result acc = 97.2%.

Stochastic computing audio synthesizer

A domain-specific language for bitstream computing

This is a transaction-level, event-driven python-based simulator for evaluation of stochastic computing based optical neural network accelerators for various quantized Convolutional Neural Network models. This can generate metrics of an accelerator like latency, area, energy consumption and power

Supplementary material for a manuscript in IEEE TCAD

Solving stochastic problems involving Markov Chains

Seminar report and presentation slides on topic Stochastic Computational Deep Belief Network

stochastic computing with dianet