Parallel Simulation of Equation-based Object-Oriented Models with Quantized State Systems on a GPU

Abstract

This work focuses on the use of parallel hardware to improve the simulation speed of equation-based object-oriented Modelica models. With this intention, a method has been developed that allows for the trans- lation of a restricted class of Modelica models to par- allel simulation code, targeted for the Nvidia Tesla architecture and based on the Quantized State Sys- tems (QSS) simulation algorithm. The OpenModel- ica Compiler (OMC) has been extended with a new back-end module for automatic generation of the sim- ulation code that uses the CUDA extensions to the C language to be executable with a General Purpose Graphic Processing Unit (GPGPU). Preliminary per- formance measurments of a small example model have been done on the Tesla architecture.