Optimization problem: bulk Boolean operations on vectors

[Observer via Digitalmars-d]

On Friday, 23 December 2016 at 16:15:44 UTC, Andrei Alexandrescu 
wrote:
> An interesting problem to look at:
>
> https://github.com/dlang/dmd/pull/6352

With respect to bulk operations on vectors, of course I recognize 
the
desire to use high-level code which is portable across platforms.
But I also wonder whether this is the sort of thing that is really
best handled these days by using the multi-media instruction set
instead of the general-purpose instruction set.  There's a lot of
special-purpose arithmetic that is provided by the CPU in such a
context, and it seems a bit silly to ignore that possibility.

Also more generally regarding low-level stuff like this, perhaps 
this
reference might be of interest:
https://www.amazon.com/Hackers-Delight-2nd-Henry-Warren/dp/0321842685