AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |
Back to Blog
Real Time Gpu Cloth Simulation4/22/2021
In SPH-based fluid simulation, the physical attributes of one particle are typically integrated over its dynamically changing neighbors at every time step, and neighbor search is the computationally most expensive part of the simulation, especially when PCISPH is used since this algorithm requires searching neighboring particles multiple times per simulation step due to its iterative feature.
![]() ![]() Moreover, we discuss several optimization techniques including using fast on-chip shared memory to avoid global memory bandwidth limitations and thus further improve performance on modern GPU hardware. With our framework, the realism of real-time fluid simulation is significantly improved since our method enforces incompressibility constraint which is typically ignored due to efficiency reason in previous GPU-based SPH methods. The performance results illustrate that our approach can efficiently simulate realistic incompressible fluid in real time and results in a speed-up factor of up to 23 on a high-end NVIDIA GPU in comparison to single-threaded CPU-based implementation. Introduction Physically based simulation methods have revolutionized special effects in computer games and interactive applications, creating incredible scenes consisting of fluids, cloth, rigid bodies, and others. Among these methods, smoothed particle hydrodynamics (SPH) is a well-known particle-based approach with a wide application for simulating many kinds of phenomena including water, fire, and gas. This paper focuses on simulating incompressible fluids for real-time visual effects. Many studies have been devoted to real-time fluid simulation methods in recent years. However, these techniques suffer from visible compression artifacts. Although weakly compressible SPH (WCSPH) is the most widespread technique for pressure conservation in SPH based methods, the challenge with WCSPH is that to get sufficient realism requires very small time step. To solve this problem, PCISPH is introduced to enforce incompressibility with significantly larger time steps. ![]() Although this makes it a good candidate for simulating realistic fluid phenomenon, it is still too computationally intensive for the state-of-the-art CPU to put into real-time applications. The modern GPU substantially outpaces its CPU counterpart in arithmetic throughput and memory bandwidth, making it the ideal processor to boost a vast variety of data parallel applications such as SPH-based fluid simulations. Originally GPU was designed for real-time graphics rendering applications. In the past decade, it has been established as one of the major parallel processors for computationally intensive tasks. Advances in massively parallel GPU architectures have opened doors to employing PCISPH-based methods for simulating realistic fluid phenomena in real time. Modern GPUs normally have thousands of compute cores which deliver tremendous computing power. PCISPH exposes a high degree of parallelism, making it a perfect candidate for state-of-the-art GPUs. In the SPH literature, several GPU-accelerated techniques have been proposed to speed up the time-consuming physics calculation. However, efficient GPU-based PCISPH solver for real-time incompressible fluids has not been presented.
0 Comments
Read More
Leave a Reply. |