Software Engineering · Performance
Actually running multiple tasks at the exact same instant, across multiple CPU cores or machines simultaneously.
Example: A video renderer splitting a frame into 8 tiles and processing all 8 simultaneously on 8 separate cores.
In practice: Parallelism buys real speedup but only for work that can be split into independent pieces — tasks with sequential dependencies between steps can't be parallelized no matter how many cores you add.