Using GPUs for Realtime Prediction of Optical Forces on Microsphere Ensembles

Sujal Bista, Sagar Chowdhury, Satyandra K. Gupta and Amitabh Varshney
Laser beams can be used to create optical traps that can hold and transport small particles. Optical trapping has been used in a number of applications ranging from prototyping at the microscale to biological cell manipulation. Successfully using optical tweezers requires predicting optical forces on the particle being trapped and transported. Reasonably accurate theory and computational models exist for predicting optical forces on a single particle in the close vicinity of a Gaussian laser beam. However, in practice the workspace includes multiple particles that are manipulated using individual optical traps. It has been experimentally shown that the presence of a particle can cast a shadow on a nearby particle and hence affect the optical forces acting on it. Computing optical forces in the presence of shadows in real-time is not feasible on CPUs. In this paper, we introduce a ray-tracing-based application optimized for GPUs to calculate forces exerted by the laser beams on microparticle ensembles in an optical tweezers system. When evaluating the force exerted by a laser beam on 32 interacting particles, our GPU-based approach is able to get a 66-fold speed up compared to a single core CPU implementation of traditional Ashkin's approach and a 10-fold speedup over the single core CPU-based implementation of our approach.
DOI

Speeding Up Particle Trajectory Simulations Under Moving Force Fields using Graphic Processing Units

Robert Patro, John P. Dickerson, Sujal Bista, Satyandra K. Gupta, Amitabh Varshney

In this paper, we introduce a graphic processing unit (GPU)-based framework for simulating particle trajectories under both static and dynamic force fields. By exploiting the highly parallel nature of the problem and making efficient use of the available hardware, our simulator exhibits a significant speedup over its CPU-based analog. We apply our framework to a specific experimental simulation: the computation of trapping probabilities associated with micron-sized silica beads in optical trapping workbenches. When evaluating large numbers of trajectories (4096), we see approximately a 356 times speedup of the GPU-based simulator over its CPU-based counterpart.

DOI

Stochastic Simulations With Graphics Hardware: Characterization of Accuracy and Performance

Arvind Balijepalli, Thomas W. LeBrun, Satyandra K. Gupta

Methods to implement stochastic simulations on the graphics processing unit (GPU) have been developed. These algorithms are used in a simulationof microassembly and nanoassembly with optical tweezers, but are alsodirectly compatible with simulations of a wide variety of assemblytechniques using either electrophoretic, magnetic, or other trapping techniques. Significant speedup is possible for stochastic particle simulations when using the GPU, included in most personal computers (PCs), rather than the central processing unit (CPU) that handles most calculations. However, a careful analysis of the accuracy and precision when using the GPU in stochastic simulations is lacking and is addressed here. A stochastic simulation for spherical particles has been developed andmapped onto stages of the GPU hardware that provide the best performance. The results from the CPU and GPU implementation are then compared with each other and with well-established theory. The error in the mean ensemble energy and the diffusion constant is measured for both the CPU and the GPU implementations. The time taken to complete several simulation experiments on each platform has also been measured and the speedup attained by the GPU is then calculated.

DOI

Generating Simplified Trapping Probability Models From Simulation of Optical Tweezers System

Banerjee AG, Balijepalli A, Gupta SK, LeBrun TW

This paper presents a radial basis function based approach to generatesimplified models to estimate the trapping probability in optical trappingexperiments using offline simulations. The difference form of Langevin's equation is used to perform physically accurate simulations of a particleunder the influence of a trapping potential and is used to estimate trapping probabilities at discrete points in the parameter space. Gaussian radial basis functions combined with kd-tree based partitioning of the parameter space are then used to generate simplified models of trapping probability. We show that the proposed approach is computationally efficient in estimating the trapping probability and that the estimated probability using the simplified models is sufficiently close to the probability estimates from offline simulation data.

DOI