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

Sure, but you could run several searches in parallel on a SIMD architecture with this.


You could always run several searches in parallel with normal binary searches as well, since this offers no substantial advantages over a branching method.

Unless I'm missing something major?


How would you run several searches in parallel on a single core without doing something like this?




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

Search: