On Thu, Jan 24, 2019 at 1:04 PM Anton Youdkevitch
wrote:
>
> Richard,
>
> Thanks a lot for the response! I will definitely
> try the constructor approach.
>
> You mentioned non-grouped store. Is the handling
> of such stores actually there and I just missed
> it? It was a big hassle for me as I di
Richard,
Thanks a lot for the response! I will definitely
try the constructor approach.
You mentioned non-grouped store. Is the handling
of such stores actually there and I just missed
it? It was a big hassle for me as I didn't manage
to find it and assumed there isn't one.
I have a question (a
On Mon, Jan 21, 2019 at 2:20 PM Anton Youdkevitch
wrote:
>
> Here is the prototype for doing vectorized reduction
> using SLP approach. I would appreciate feedback if this
> is a feasible approach and if overall the direction is
> right.
>
> The idea is to vectorize reduction like this
>
> S = A[0
