Re: [lmi] Using auto-vectorization

[Greg Chicares]

On 2017-01-24 02:49, Vadim Zeitlin wrote:
[...]
> ET-based code seems to profit from auto-vectorization just
> as well as everything else, so I don't see any reason to use anything else,
> especially if the code clarity and simplicity are the most important
> criteria.
> 
>  Now, whether using the particular PETE library is the best choice in 2017
> is another question and I suspect that it isn't, but I'm not aware of any
> critical problems with it neither.

It seems that there was a flurry of interest around the turn of the
century, but almost none since then. The audience for ET libraries is
relatively small, and I'd guess that most potential users chose a
library long ago and aren't interested in changing.

>  So, I guess, I'm still not sure what, if anything, should be done here? I
> can spend a lot of time profiling/benchmarking/debugging and it probably
> will result in at least some useful insights, but I can't propose any
> syntax better than the current ET-based one and so I'm still not sure what
> is my goal here.

I think we're done for now. We aren't likely to find anything that
outperforms PETE. We can make greater use of it as time permits.