Scaling of run time with the number of processors. Speed-up was calculated as the run time for the batch job on one processor divided by he run time with the given number of processors. (A) Speed-up for the two gene (geneA_geneB_CFTT_1p1) exemplar model. (B) Speed-up for the one gene (geneA_CFTT_0p0) model with 1× and 10× parameter sets. All models were run 10,000 times as a batch job using the BNS software on a HP XC machine with distributed memory architecture using up to 100 processors.