An Investigation of Parallel Programming Techniques Applied to Monte Carlo Simulations for Post-Flight Reconstruction of Spacecraft Trajectory

Parallelizing software to execute on multi-core central processing units (CPUs) and graphics processing units (GPUs) can be challenging. For some fields outside of Computer Science, this transition comes with new issues. For example, memory limitations can require modifications to code not initially developed to run on GPUs. This work applies the Open Multi-Processing (OpenMP) and Open Accelerators (OpenACC) directive-based parallelization strategies on a Monte Carlo simulation approach for trajectory reconstruction enabling it to run on multi-core CPUs and GPUs. Large matrix operations are the most common use of GPUs, which are not present in this algorithm; however, the natural parallelism of independent trajectories in Monte Carlo simulations is exploited. Benchmarking data are presented comparing execution times of the software for single-thread CPUs, multi-thread CPUs with OpenMP, and multi-thread GPUs using OpenACC. These data were collected using nodes with Intel ® Xeon ® E5-2670 (Sandy Bridge) CPUs enhanced with NVIDIA ® Tesla ® K40 GPUs on the Pleiades Supercomputer cluster at the National Aeronautics and Space Administration (NASA) Ames Research Center (ARC) and a local Intel ® Xeon Phi ™ node at NASA Langley Research Center (LaRC). and orientation), and integrates the inertial measurement unit (IMU) data to determine the vehicle states throughout its flight. Lugo et al. 1 developed a Monte Carlo based approach for trajectory reconstruction that incorporated the vehicle’s final state information and introduces statistics. This method decreases uncertainties in the reconstruction results, which improves model validations and post-flight analysis. However, this Monte Carlo approach requires the integration of several thousand trajectories. These calculations are time consuming when executed serially, but the execution time can be decreased by utilizing concurrent computation. This paper examines the use of parallel programming techniques on an algorithm that applies inertial navigation to trajectory reconstruction in a Monte Carlo dispersion process. The two parallel programming techniques being utilized are OpenMP and OpenACC, which are used on multi-core CPUs and GPUs, respectively. Two studies are conducted to determine optimal performance based on thread count with OpenMP and register per thread for OpenACC. Additionally, comparisons are shown between three different compilers and three different types of hardware. or V100, will tested in future work.