On Saturday, June 25, 2016 6:45:05 AM CDT jw3126 wrote:
> @Tim: Thanks for the insights! On my machine slowness starts to kick in at 
> size 9 already.

Depends on which version of julia you're running (I'm running a recent 
julia-0.5-dev).

> I tried to read the llvm code, but did not really 
> understand it. It seems however that the machine will not go through N 
> (out, t) pairs for a tuple of length N?

The @inline gives LLVM the chance to elide the apparent intermediates: 
essentially you're describing the computation you want to perform in a way the 
compiler can understand, but which may or may not reflect the final operations 
performed by the CPU. Coaching it into taking the opportunity to fully 
optimize this operation takes a little practice, and unfortunately sometimes 
involves a little magic (e.g., https://github.com/JuliaLang/julia/issues/
17126).

Best,
--Tim

Reply via email to