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

Except that actually does apply to X86.

As an example with AVX instructions you can process 8 floats at the same time, compared to 4 doubles. So if float is enough for you you can expect double performance in either memory transfer bound or ideally vectorizable algorithms.

And in mobile computer graphics 16bit values are common.



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

Search: