Multithreading broadcasted array ops

[pxshen]

What’s best practice for multithreading broadcasted array ops?
I know of various multithreaded map functions but wish for a simple macro like @threads A.+B doesn’t seem to work?
Even better why can’t compiler automatically multithread broadcasted calls if Julia is started w/ multithreads?

[jar1]

I like to use Tullio.jl or OhMyThreads.jl for simple multithreading tasks.

[rveltz]

@tturbo

[pxshen]

docs says @tturbo has yet to support broadcasting?

[stevengj]

@pxshen

I know of various multithreaded map functions but wish for a simple macro like @threads A.+B doesn’t seem to work?

multi-threaded (@threads) dotcall/broadcast?

Multithreaded broadcast?

So, with the new improvements to threading, I’m wondering how that can be exposed easily to users. One of the most common parallelizable pattern is broadcasting (discussed in https://github.com/JuliaLang/julia/issues/19777). The conservative approach is to make @threads work for these, so we can do @threads a.= b.+ c. This is fine, but adding annotations gets annoying really fast (“so, what combination of @threads @simd @inbounds @. do I need this time?”), especially in matlab/numpy vector code …

What is the “standard” way to use multi-threading with dot-calls (vectorized/broadcast)?

All the discussions I’ve found on this topic are years old: “Multithreaded broadcast?” thread use multiple threads in vectorized operations · Issue #1802 · JuliaLang/julia · GitHub multi-threaded (@threads) dotcall/broadcast? · Issue #19777 · JuliaLang/julia · GitHub `@threads f(...)` syntax to tell callee to use threaded version? · Issue #40329 · JuliaLang/julia · GitHub What is currently the most straight-forward way to get multi-threading with operations like these: a = rand(4000, 4000) b…

GitHub - Jutho/Strided.jl: A Julia package for strided array views and...