The title of this is misleading, and so is this part: >>The cost of a single increment operation is highly variable. At low boundary values, it is around 0.03

The title of this is misleading, and so is this part: >>The cost of a single increment operation is highly variable. At low boundary values, it is around 0.03 time units per increment. But at high boundary values, the cost drops to one tenth that. The cost of an increment operation has nothing to do with it. It's very fast and doesn't change. What matters is the cost of flushing the pipeline on every mispredicted branch. Maybe you're supposed to get that that part is "naive" and ignore it once you've read the whole article?