On: Optimizing Linux Memory Management for Low-latency...

[erichocean]

I understood it to mean that it had 1/4 the error rates they were previously seeing.

[dllthomas → erichocean]

Huh, that's not even one I considered. I guessed 1/5, so the change was 4x the final value. Either way it's wrong - "dropped by X%" should be calculated as (original - final) / original.

[apurvamehta → erichocean]

This is exactly right :)

[caf → apurvamehta]

Then it dropped by 75%.

[apurvamehta → caf]

Yes. It was a blunder. The post has been updated to reflect this.