Final answer:
Message passing is useful for parallel applications that degrade when parts are not running.
Step-by-step explanation:
Message passing is useful for medium-grained to fine-grained parallel applications whose performance severely degrades when any part of the application is not running while other parts are ready to run. In message passing, different parts of the application communicate by sending messages to each other. This ensures that each part can operate independently without waiting for other parts to be ready.
For example, in a distributed computing system, where multiple computers work together to solve a problem, message passing allows different computers to exchange data and coordinate their actions to achieve parallel processing. If any computer is not running, other computers can still continue with their tasks and make progress.
Synchronisation in parallel computing is useful for medium-grained to fine-grained parallel applications whose performance severely degrades when any part of the application is not running while other parts are ready to run. These applications typically require that processing elements work in a coordinated manner, often relying on mechanisms such as barriers, locks, and semaphores to ensure all tasks progress in step with each other.
Failing to do so can result in some processes stalling while waiting for others, which can lead to inefficiencies and performance bottlenecks.