Abstract
AbstractRecently the hardware performance of mobile devices have been extremely increased and advanced mobile devices provide multi-cores and high clock speed. In addition, mobile devices have advantages in mobility and portability compared with PC and Console, so many games and simulation programs have been developed under mobile environments. Physically-based simulation is a one of the key issues for deformable object modeling which is widely used to represent the realistic expression of 3D soft objects with tetrahedrons for game and 3D simulation. However, it requires high computation power to plausibly and realistically represent the physical behaviors and interactions of deformable objects. In this paper, we implemented parallel cloth and mass-spring simulation using graphics processing unit (GPU) with OpenCL and multi-threaded central processing unit (CPU) on a mobile device. We applied CPU and GPU parallel computing technique into spring force computation and integration methods such as Euler, Midpoint, 4th-order Runge-Kutta to optimize the computational burden of dynamic simulation. The integration methods compute the next step of positions and velocities in each node. In this paper, we tested the performance analysis for the spring force calculation and integration method process using CPU only, multi-threaded CPU, and GPU on mobile device respectively. Our experimental results concluded that the calculation using proposed multi-threaded CPU and GPU multi-threaded CPU are much faster than using just the CPU only.
Highlights
As the hardware performance of mobile devices have been noticeably increased, new applications with utilizing the advanced performance of mobile devices have been developed and many traditional applications that were already developed based on PC environments have been converted to mobile devices [1]
We proposed and implemented the cloth simulation and 3D mass-spring simulation using multi-thread approach and Open computing language (OpenCL) library to compute node position and spring force information with central processing unit (CPU) and general-purpose computing on graphics processing unit (GPGPU) parallel computing
When the simulation data is stored in the CPU, OpenCL starts the computation of spring forces and new positions are predicted using a lot of graphics processing unit (GPU) ALUs
Summary
As the hardware performance of mobile devices have been noticeably increased, new applications with utilizing the advanced performance of mobile devices have been developed and many traditional applications that were already developed based on PC environments have been converted to mobile devices [1]. Representation of a 3D world or physically-based simulation requires lots of physical calculations, but the performance of mobile devices is not high enough to perform the numerical problems. To alleviate these problems, researchers have studied several techniques such. We provide a guideline for modeling of 3D objects and selecting of integration methods for physically based simulation on mobile device using parallel computing approach. We implemented a new data structure to calculate the spring forces in parallel computing for deformable object simulation which has uneven spring connections on each node.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
