This really needs to be upvoted harder. The post doesn't say what units the terms a measured in, but it's evident from a quick test that the unit is ms. And wh

This really needs to be upvoted harder. The post doesn't say what units the terms a measured in, but it's evident from a quick test that the unit is ms. And when you're in the sub-ms time domain, if you don't know what you're doing, benchmarking is apt to reflect noise in its results.