Package topics.parallel

package topics.parallel

Parallel algorithms using Java's Fork/Join Framework.

Parallel algorithms divide work among multiple processor cores to reduce wall-clock time. This package demonstrates parallelisation through the divide-and-conquer paradigm using ForkJoinPool and its RecursiveAction / RecursiveTask abstractions.

Key Concepts

• Work splitting — sub-problems below a threshold are solved sequentially to amortise task-creation overhead.
• Work stealing — idle threads steal tasks from busy threads' queues, maximising CPU utilisation.
• Amdahl's Law — speedup is bounded by the sequential fraction of the computation.

Classes in This Package

• FibonacciAlgorithm — sequential Fibonacci baseline
• FibonacciTask — parallel Fibonacci via RecursiveTask
• FileProcessingTask — parallel file processing
• RecursiveActionComparison — compares sequential vs parallel actions
• RecursiveActionSquare — parallel squaring via RecursiveAction
• RecursiveTaskSum — parallel array sum via RecursiveTask

Classes

Class	Description
FibonacciAlgorithm	To solve the classical Fibonacci algorithm using a recursive and inefficient algorithm
FibonacciTask	To calculate the Fibonacci of a number n
RecursiveActionComparison	To calculate the cube root of the values of an array
RecursiveActionSquare	To calculate the square of the values of an array
RecursiveTaskSum	To calculate the addition of the values of an array