Re: [boost] Accelerating algorithms with SIMD - Segmented iterators and alternatives

Subject: Re: [boost] Accelerating algorithms with SIMD - Segmented iterators and alternatives
From: joel falcou (joel.falcou_at_[hidden])
Date: 2010-10-12 15:23:03

On 12/10/10 21:15, Simonson, Lucanus J wrote:
> You can't have your cake and eat it too.  If you leave the low level optimization to the C++ compiler you have to accept that it will punt when it can't be sure whether two pointers might be aliases, for example.
>    
This can be checked quite lightweightly at runtime after some ET based 
exploration. We did it in NT2 and it works out of the box.
> I wasn't trying to blame you for anything other than dismissing the JIT approach out of hand and my only intention was to get you to spend some time looking into this CT thing which looks quite promising.
Ct and Rapidmind have been around since ages, except nobody did anythign 
with it ever and no actual publications on them where available except 
for the usual marketing-fueled whitepapers. Until now, for me, that's 
the definition of vaporware.