Accelerating Geophysics Simulation using CUDA

Brandon Holt and Daniel Ernst

Volume 2, Issue 1 (December 2011), pp. 21–27

PDF icon Download PDF

  author={Brandon Holt and Daniel Ernst},
  title={Accelerating Geophysics Simulation using CUDA},
  journal={The Journal of Computational Science Education},
Copied to clipboard!

CitcomS, a finite element code that models convection in the Earth's mantle, is used by many computational geophysicists to study the Earth's interior. In order to allow faster experiments and greater simulation capability, there is a push to increase the performance of the code to allow more computations to complete in the same amount of time. To accomplish this we leverage the massively parallel capabilities of graphics processors (GPUs), specifically those using NVIDIA's CUDA framework. We translated existing functions to run in parallel on the GPU, starting with the functions where the most computing time is spent. Running on NVIDIA Tesla GPUs, initial results show an average speedup of 1.8 that stays constant with increasing problem sizes and scales with increasing numbers of MPI processes. As more of the CitcomS code is successfully translated to CUDA, and as newer general purpose GPU frameworks like Fermi are released, we should continue to see further speedups in the future.