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

Isn't a faster way to, instead of using a backing array, to use a pool of memory that you just point into with pointers addressed with the hash?

Nasty stuff that you can do in a language like C :)



As far as the CPU is concerned, there isn't a big difference if you access a block of memory via pointers or via an array. Converting the hash result to a memory address is not the bottleneck. The main bottlenecks are the hash function itself and the logic for dealing with collusions


I definitely agree that is where the main bottlenecks are!




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

Search: