Any SIMD experts?

[John Colvin via Digitalmars-d]

On Monday, 8 December 2014 at 23:40:21 UTC, Martin Nowak wrote:
> On 12/08/2014 06:20 PM, John Colvin wrote:
>>
>> To conceptually get what it's doing here, the trick is that 
>> it's
>> offsetting the values so as to simulate unsigned comparisons 
>> using
>> signed instructions.
>
> All too easy, but would've taken me a pen and paper to realize 
> :).

I wouldn't have seen it so easily but I've spent the last few 
weeks doing discrete Fourier analysis, so I've got very used to 
spotting modulo arithmetic tricks of exactly this kind.