The state of loop optimization in Julia

We have loop vectorization which optimizes a single, highly restricted loop, then we will have loopmodel with less restriction but no multithreading. Then, there is a “future work” for loopvectorization. Naturally, one would beg to wonder to combine all the advantages of the loop libraries. Then, where do we go from here? Beat C/C++ at this and Julia will be the first language to fully optimize loops.

Where is it now?

Bold claim.
I remember seeing what intel’s fortran compiler (ifort), or maybe it was Cray’s fortran compiler?) was doing back in 2014.