Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

This is the benchmark they wrote: https://github.com/eliben/code-for-blog/blob/7278526923168d2...

The benchmark alternates between ~1 million different keys to check in the filter, explicitly to account for cache effects.



A single lookup is going to take more than 30ns, the reason they only see that is that the OoO machinery of their CPU is good enough to run those lookups in parallel.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: