Does `@distributed` return an array of futures?

General Usage
,
1

The Julia manual (Multi-processing and Distributed Computing · The Julia Language) claims that @distributed without a reduction operator returns a an array of Futures:

Using “outside” variables in parallel loops is perfectly reasonable if the variables are read-only:

a = randn(1000)
@distributed (+) for i = 1:100000
    f(a[rand(1:end)])
end

Here each iteration applies f to a randomly-chosen sample from a vector a shared by all processes.

As you could see, the reduction operator can be omitted if it is not needed. In that case, the loop executes asynchronously, i.e. it spawns independent tasks on all available workers and returns an array of Future immediately without waiting for completion. The caller can wait for the Future completions at a later point by calling fetch on them, or wait for completion at the end of the loop by prefixing it with @sync, like @sync @distributed for.

But that doesn’t seem to be the case? I ran the following on my laptop (Julia version 1.11.4):

$ julia -p 2

julia> @everywhere f(x) = 2x

julia> a = randn(1000)
       @distributed (+) for i = 1:100000
         f(a[rand(1:end)])
       end
10268.871546420913

julia> @distributed for i = 1:100000
         f(a[rand(1:end)])
       end
Task (runnable, started) @0x000077c9cf892bd0

julia> ret = @distributed for i = 1:100000
         f(a[rand(1:end)])
       end
Task (runnable, started) @0x000077c9cf23b3a0

julia> typeof(ret)
Task

It looks to me like @distributed without a reducer returns a Task, not an array of Futures. Am I misunderstanding something?

2

Looks like Task is the type of the last object in the loop. How about trying something like result[i] = f(a[rand(1:end)]) ?

Edit: In the documentation, there is an example of

a = zeros(100000)
@distributed for i = 1:100000
    a[i] = i
end

in page Multi-processing and Distributed Computing · The Julia Language

3

That also returns a task:

julia> a = zeros(100000);

julia> @distributed for i = 1:100000
           a[i] = i
       end
Task (runnable, started) @0x000077c9cf893b70