I looked at https://github.com/dg92/Performance-analysis-es6/blob/master... briefly and saw that the different algorithms always run one after another.

Does the first algorithm get an unfair cold cache disadvantage?