Maybe the title of the post should be modified accordingly : “Julia is 500 % faster than SAS We need slower Julia please (it gets confusing)”.
I hope the OP will forgive me this bad joke, but I wonder how many threads claiming that “Julia was slower than XXX” I have seen so far on discourse. Almost always with the same conclusion…
Disclaimer : “I have already wrote a similarly wrong thread Sum performance for Array{Float64,2} elements” 