[nudpiedo]

In this case it does by hitting hardware maximum. It scales until the max it is possible, due the bottleneck being in the hardware, not in the database or i/o. Check the article again:

> At maximum throughput, Postgres CPU is fully utilized, showing the database is actually saturated instead of bottlenecked on contention.

[orf]

It doesn’t necessarily follow that because Postgres is “fully utilised” it is hitting “hardware maximum”, or even that it’s not bottlenecked on contention.

As an example, spinlocks can push CPU usage very high whilst being an obvious symptom of contention.