Skip to main content


Fig. 17 | BMC Systems Biology

Fig. 17

From: Optimisation of an exemplar oculomotor model using multi-objective genetic algorithms executed on a GPU-CPU combination

Fig. 17

Speedup obtained with the parallel NSGA-II method versus the number of parallel instances using the AMD Firepro W8100 GPU. Speedup is defined as the ratio of parallel execution time to serial execution time for the same number of NSGA-II instances. The indexed numbers indicate the population size for each NSGA-II instance. In each case, the sum of the populations of all NSGA-IIs running in parallel is 32000; this is in the range of parallel integrations that provides the maximum speedup, compared to the CPU (see Fig. 16)

Back to article page