Re: Implementation of pop count instruction

Oliver S. wrote:

> Can anyone tell me for what this popcnt-instruction should be good
> for? I'm able to imagine some useful cases for such an instruction
> but I can't imagine where such an instruction would give a perfor-
> mance-boost of a complete algorithm.

Detecting non-random (i.e. skewed) output from a trial decrypt operation?

'No Such Agency' is rumoured to be the main reason for popcnt on most
architectures where it has been included.

Terje

--
- <Terje.Mathisen@xxxxxxxxxxxxx>
"almost all programming can be viewed as an exercise in caching"
.

Relevant Pages

  • Re: Implementation of pop count instruction
    ... >> Can anyone tell me for what this popcnt-instruction should be good ... I'm able to imagine some useful cases for such an instruction ... > => having a fast popcnt allows fast paths for common computations ...
    (comp.arch)
  • Re: Implementation of pop count instruction
    ... | Can anyone tell me for what this popcnt-instruction should be good ... I'm able to imagine some useful cases for such an instruction ... Prev by Date: ...
    (comp.arch)
  • Re: Implementation of pop count instruction
    ... I'm able to imagine some useful cases for such an instruction but I can't imagine where such an instruction would give a perfor- mance-boost of a complete algorithm. ... A google search will provide ... A popular claim is that this instruction is ...
    (comp.arch)
  • Re: FA: Swiss Incabloc System Model
    ... cc READER wrote: ... I imagine it could have been used for ... instruction, at trade shows and even as a sales tool for demonstrating ...
    (alt.horology)
  • Re: FA: Swiss Incabloc System Model
    ... I imagine it could have been used for ... instruction, at trade shows and even as a sales tool for demonstrating ...
    (alt.horology)