FPGA of Acceleration of Stochastic Simulation

By Tian Gao

This project is designed to implement a stochastic algorithm on an FPGA system to accelerate a biological simulation. A fast national disease spread prediction algorithm is needed to anticipate the tendency of an infectious disease to spread in a given geography. Using a traditional computing system, the simulation would require large execution time as the model size grows. The project explores methods for implementing such an algorithm on an FPGAsystem. By utilizing the parallel features of an FPGA, we may find a way to run the whole simulation in real time. In this project, a MATLAB simulation is firstly designed for a hardware compatible algorithm for an SIR model, which is a basic disease spread model. Then, three kinds of FPGAs are used to implement the algorithm. Finally, some verification is applied to validate the result. The project has successfully implemented a hardware compatible algorithm for an SIR model on an FPGA system.

Full report (pdf)

Poster (pdf)