Fat vs. thin threading approach on GPUs: application to stochastic simulation of chemical reactions

Fat vs. thin threading approach on GPUs: application to stochastic simulation of chemical reactions

Tue, 19/10/2010
13:15 		Guido Klingbeil (Centre for Mathematical Biology) Junior Applied Mathematics Seminar Add to calendar DH 1st floor SR
Fat vs. thin threading approach on GPUs: application to stochastic simulation of chemical reactions
We explore two different threading approaches on a graphics processing unit (GPU) exploiting two different characteristics of the current GPU architecture. The fat thread approach tries to minimise data access time by relying on shared memory and registers potentially sacrificing parallelism. The thin thread approach maximises parallelism and tries to hide access latencies. We apply these two approaches to the parallel stochastic simulation of chemical reaction systems using the stochastic simulation algorithm (SSA) by Gillespie. In these cases, the proposed thin thread approach shows comparable performance while eliminating the limitation of the reaction system's size.Link to paper:  http://people.maths.ox.ac.uk/erban/papers/paperCUDA.pdf