It's a microbenchmark (if even that), take it with a grain of salt. You'd probably see a bigger difference with bigger/more/more complicated tasks,