What is the difference between parallelism and concurrency?

Parallelism and concurrency are two related but different concepts in programming and systems.

Concurrency is the ability of a system to manage multiple tasks that can be executed concurrently in time, but not necessarily at the same instant. Tasks can switch, creating the illusion of simultaneous execution. For example, a single-threaded application with asynchronous input/output.

Parallelism is the simultaneous execution of multiple tasks at the same moment in time, usually on different processors or cores. This is physical simultaneous execution.

In other words, concurrency is about the structure and design of a program that allows working with multiple tasks, while parallelism is about the actual simultaneous execution of these tasks.

Example:

• In a single-threaded application with concurrency, tasks switch over time.
• In a multi-processor system with parallelism, tasks are executed simultaneously on different cores.