The benchmark with 1B rows in this blogpost seems irrelevant for comparing performance of different programming languages.

It seems like the execution time of a program would be dominated by loading data from the file. And a lot of people posted solution with specs of cpu but not specs of disk (hdd, ssd, raid) although that seems more relevant.

Why would they compare languages and solutions in this way?