Introduction to Parallel Computing with Elixir

Elixir is presented as a language for real-time software because it runs on Erlang’s virtual machine, reuses Erlang libraries, and adds syntax, metaprogramming, and fault-tolerance tools. The post first builds a Fibonacci example with Mix and then turns it into a parallel version using spawned processes and message passing so results can be gathered and ordered.

It also shows how to package the project as an escript by adding a main function and configuring Mix, which makes the tool runnable from the command line without installing Elixir on the target machine.

Reading notes

#

• Elixir is described as compiled, built on Erlang’s virtual machine, and able to use Erlang libraries.
• The language is introduced as useful for concurrency, parallel computing, fault tolerance, and real-time software.
• A new Mix project is created with mix new fibex.
• A Fibonacci module is implemented using tail recursion, public functions with def, private functions with defp, and the |> operator.
• The sequential version is run from the terminal with mix run and returns the expected Fibonacci results.
• The parallel version uses spawn to run Fibonacci work in new lightweight processes.
• Message passing is used to return results when they are ready.
• The code carries each input’s index so the results can be sorted after concurrent execution.
• The code also carries the input length and decrements it in the receive loop to know when computation is finished.
• An escript is built by adding a main function and configuring mix.exs.
• The command-line executable is built with mix escript.build and run with numeric arguments.
• The summary says the result is a command-line utility that uses multiple CPUs efficiently and points readers to deeper Elixir and Erlang study.