you're missing the point. Postgres performs well on large CPU. Postgres as-used by OP does not and is a waste of money. It's great that he benchmarked for a larger CPU, that's not what people are disputing, they are disputing the ridiculous conclusion.