Fig. 5

The bottom row depicts the model execution timings on GPU when using three different amounts of age categories. The x-axis depicts the amount of GPU threads used when evaluating an ODE, while the y-axis depicts the execution timing in milliseconds. The number of operations is in the order of \(\mathcal {O}(n^{2})\), meaning that when comparing 50 and 100 age categories, the latter has 4 times as many operations. One can see that when there are more age categories available, the better the GPU will perform when compared to the CPU. The top row depicts the speedup of the GPU when compared with the best CPU timing for each of the three age category settings