Eugenio Rustico

Advances in Multi-GPU Smoothed Particle Hydrodynamics Simulations

We present a multi-GPU version of GPUSPH, a CUDA implementation of fluid-dynamics models based on the smoothed particle hydrodynamics (SPH) numerical method. The SPH is a well-known Lagrangian model for the simulation of free-surface fluid flows; it exposes a high degree of parallelism and has already been successfully ported to GPU. We extend the GPU-based simulator to run simulations on multiple GPUs simultaneously, to obtain a gain in speed and overcome the memory limitations of using a single device. The computational domain is spatially split with minimal overlapping and shared volume slices are updated at every iteration of the simulation. Data transfers are asynchronous with computations, thus completely covering the overhead introduced by slice exchange. A simple yet effective load balancing policy preserves the performance in case of unbalanced simulations due to asymmetric fluid topologies. The obtained speedup factor (up to 4.5x for 6 GPUs) closely follows the expected one (5x for 6 GPUs) and it is possible to run simulations with a higher number of particles than would fit on a single device. We use the Karp-Flatt metric to formally estimate the overall efficiency of the parallelization.

Smoothed Particle Hydrodynamics Simulations on Multi-GPU Systems

We present a multi-GPU version of GPUSPH, a CUDA implementation of fluid-dynamics models based on the Smoothed Particle Hydrodynamics (SPH) numerical method. SPH is a well-known lagrangian model to simulate fluid flows, it exposes a high degree of parallelism and has already been successfully ported to GPU. We extend the GPU-based simulator to run simulations on multiple GPUs simultaneously, to obtain a gain in speed and overcome the memory limitations of using a single device. The computational domain is spatially split with minimal overlapping and shared volume slices are updated at every iteration of the simulation. Data transfers are asynchronous with computations, thus completely covering the overhead introduced by slice exchange. The obtained speedup factor differs from the ideal one by a small cost function linear in the number of devices, and it is possible to run simulations with a higher number of particles than would fit on a single device.

Simulation of Nearshore Tsunami Breaking by Smoothed Particle Hydrodynamics Method

AbstractThis study applies the numerical model GPUSPH, an implementation of the weakly compressible smoothed particle hydrodynamics (SPH) method on graphics processing units, to simulate nearshore tsunami processes. Two sets of laboratory experiments that involve violent wave breaking are simulated by the three-dimensional numerical model. The first set of experiments addresses tsunamilike solitary wave breaking on and overtopping an impermeable seawall. Comparison with free-surface profiles and laboratory images shows that GPUSPH satisfactorily reproduces the complicated wave processes involving wave plunging, collapsing, splash-up, and overtopping. The other set of experiments investigates tsunamilike solitary wave breaking and inundation over shallow water reefs. The performance of GPUSPH is evaluated by comparing its results with (1) experimental data including free-surface measurements and cross-shore velocity profiles, and (2) published numerical results obtained in two mesh-based wave models: the n...

Low cost finger tracking on flat surfaces

We present a flexible system to track the movements of a bare finger on a flat surface. The proposed system is able to discriminate whether the user is touching or just pointing at the surface. The system works using two webcams and a fast scanline-based algorithm. The initial setup of the two webcams is easy and fast. No markers, gloves, or other hand-held devices are required. Since the system is independent from the nature of the pointing surface, it is possible to use a screen or a projected wall as a virtual touchscreen. The complexity of the algorithms used by the system grows less than linearly with resolution, making the software layer very lightweight and suitable also for low-powered devices like embedded controllers.

Full 3D Numerical Simulation and Validation of a Fish Pass with GPUSPH

We present a validated fully three-dimensional simulation of a vertical-slot fish pass with GPUSPH, a high-performance CUDA implementation of the Smoothed Particles Hydrodynamics (SPH) numerical method for free-surface flows. The GPUSPH results are compared to flow velocity and water level measurement from a laboratory model with the same geometry. The results show good agreement between the numerical simulations and the experimental data.