I would like to see the same results with, say, 1,000 trials and 10,000 trials as well—how much of the slowdown is a fixed overhead, and how much of it is that

I would like to see the same results with, say, 1,000 trials and 10,000 trials as well—how much of the slowdown is a fixed overhead, and how much of it is that the code runs X-times slower?