Particle to Grid Transfer optimization on Parallel implementation of MPM based Fluid Simulation

• Main Author: Bimawan S, Hendrikus
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/36964
• Institution: Institut Teknologi Bandung

Fluid Simulation has uses in many subjects such as aerodynamics analysis, weather simulation, dan engine analysis. A lot of optimization has been done to improve the performance of a fluid simulation. One of them is to implement parallelization within said simulation. One of the ways to produce a fluid simulation is through implementation of Material Point Method (MPM). In the previous works MPM has been successfully implemented on Graphics Processing unit (GPU) using Compute Unified Device Architecture (CUDA). That implementation goal is to run the simulation within parallel code. This paper discusses a way to improve the performance of said simulation further through the implementation of sorting algorithm. This implementation speeds up most of the kernel performance, though an overhead for sort is introduced. The end result may improve the simulation performance for a machine configuration and may decrease the performance for others.