Please does the “Threads.@threads for” functionality split the loop into threads on RPi (v. 4)? I changed the number of threads to 4 via the env JULIA_NUM_THREADS=4, confirmed by Threads.nthreads(). Yet top command shows only one core is running in such loop. On amd64 machines the threading works great.
function generateSine(x) Threads.@threads for i=1:4 sin.(x); end end cnt = Int64(1e7); x = rand(cnt); GC.enable(false); @time generateSine(x) GC.enable(true);
Thanks a lot for any info/hints.