Key words and concepts

From Parawiki

Definitions

Parallel computing refers to solving a task faster by employing multiple processors simultaneously. Distributed computing uses a collection of autonomous computers that are interconnected with each other and cooperate, sharing resources. Note that parallel computing and distributed computing are different!
There are many parallel computing-related notions, but they are all different.

• Concurrent computing
• Highly parallel computing
• High performance computing
• Networked computing
• Decentralized computing
• Collaborative computing

A programming environment (PE) is an environment that basically provides all the tools needed to design, code, and debug parallel and distributed applications according to a given programming model or language. A problem solving environment (PSE) provides a set of user-friendly mechanisms and tools that allow to build-up an application within a specific application domain, by gluing together, with an intuitive compositional model and using some kind of problem-oriented language, different building blocks.

Ref. P. D'Ambra et al. Advanced Environments for parallel and distributed applications: a view of current status Parallel Computing 28 (2002) 1637-1662

Reasons of parallel / distributed computing

There are several reasons for deploying parallel and distributed computing:

• Absolute performance
• Price/Performance ratio
• Technological reasons
• Inherently parallel or distributed applications
• Resource sharing
• Availability
• Incremental growth