alt Hacker NewsEssentially for loop of 10k iterations comparing `fn()` versus `await fn()` fed into a microbenchmark tool, with some fiddling to detect if elimination was happening or ordering was changing things.
I was bumping into PRs trying to eliminate awaits in long loops and thinking surely the overhead can’t be so high to warrant doing this, especially after node ~16. I was wrong.
Clarifying:
‘fn()’ is the exact same function call in both cases. Only the benchmark itself was async versus sync.
Because the case under test is what’s the cost of making an async api when 90% of the calls are sync? And the answer is a lot higher than I thought.