Parallelism occurs when a computer with multiple CPUs or cores splits tasks across these components, achieving tremendous speed boosts by executing multiple operations simultaneously.
This is fundamentally different from Concurrency, which involves interleaving task execution on a single core. While concurrency is about dealing with multiple things at once through task switching, parallelism is about doing multiple things at once through simultaneous execution.
In Ruby Concurrency Mechanisms, traditional threads provide only concurrency due to the Global Interpreter Lock, which ensures only one thread executes Ruby code at a time. True parallelism requires Ractors (introduced in Ruby 3.0), where each Ractor maintains its own GIL, allowing multiple Ractors to execute Ruby code simultaneously across multiple CPU cores.
The distinction matters for performance: parallelism provides actual speedup for CPU-bound tasks, while concurrency optimizes wait time for O-bound tasks.