There are different levels of parallelism to consider, such as bit level, instruction level, data and task parallelism. Different types of parallelism have varied purposes, depending on whether they are multi-core, multi-processor, stand-alone or networked computers. Parallelism is often confused with clusters and grid computers, which are multiple computers designed to work on the same task -- as opposed to sharing individual computer resources, such as processor or memory.
1). Identify the hardware attributes of your computers. Take an inventory of all the pieces of equipment, including computers, that you will possibly be using in your parallel configuration. Note all the attributes of a particular device, include its processing power, type, disk capacity and memory.
2). Choose the levels of parallelism required for your situation. The parallelism may operate at more than one level, including bit-level, instruction level, data, or be task-oriented.
