4 ms·
The article talks about IO, but then benchmarks sha256. The node program doesn't work well because he's synchronously doing this: for (var i = 0; i < n; i++
by jwlake 9y ago
The article talks about IO, but then benchmarks sha256. The node program doesn't work well because he's synchronously doing this:
for (var i = 0; i < n; i++) { sha256(data); }
He should be using crypto and the async versions. This benchmark actually is just measuring the speed of your sha256 implementation, which I would guess is equal on all 4 platforms if you actually do them correctly.
- styfle 9y agoWhere did you even find the source code?
- mvkg 9y agohttps://peabody.io/post/server-env-benchmarks/ https://peabody.io/post/server-env-benchmarks/
- fleitz 9y agoYeah, if you are running node in production on a multicore machine you want to run multiple node instances... I do find it hilarious that the article talks IO and then benchmarks a CPU bound app.