Bone structure analysis on multiple GPGPUs

Abstract

Osteoporosis is a disease that aects a growing number of people by increasing the fragility of their bones. To improve the understanding of the bone quality, large scale computer simulations are applied. A fast, scalable and memory ecient solver for such problems is ParOSol. It uses the preconditioned conjugate gradient algorithm with a multigrid preconditioner. A modication of ParOSol is presented that prots from the exorbitant compute capabilities of recent generalpurpose graphics processing units (GPGPUs). Adaptations of data structures for the GPGPU are discussed. The fastest implementation on a GPGPU achieves a speedup of more than ve compared with the CPU implementation and scales from 1 to at least 256 GPGPUs.