[PATCH] D25722: Improved cost model for FDIV and FSQRT

[Michael Kuperstein via llvm-commits]

mkuper added a comment.

In https://reviews.llvm.org/D25722#578712, @RKSimon wrote:

> In https://reviews.llvm.org/D25722#578664, @mkuper wrote:
>
> > I'd really like to understand the discrepancy between Agner's numbers and the IACA numbers - especially since IACA does give the expected number for the latency.
> >  Andrew, any chance you could ping the IACA people at Intel and ask them about this?
>
>
> IACA doesn't have that great support these days - the website says development has been suspended, although they have announced they will do a BDW/SKL bugfix release at some point. Similar comments about dodgy costs don't seem to be answered.


To the best of my knowledge, Andrew works for Intel, so he may have a better chance of getting an answer. :-) 
As to support - the website has a comment from July that states that  "[they] are resuming support for Intel(R) Architecture Code Analyzer with BDW and SKL support probably before end of 2016".

So, one can hope...


https://reviews.llvm.org/D25722