We can learn quite a bit about the use and design of parallel computers by studying parallelism in nature and society. In fact, one can view society or culture as a set of rules and conventions to allow people to work together, i.e., in parallel, effectively, and harmoniously.
A simple illustration is the way we tackle a large project---the construction of the space shuttle. It would be attractive to solve this sequentially by hiring a single superman to complete this project. This is prohibited by current physical phenomenology, and so instead one puts together a team, maybe in this case involving 100,000 ``ordinary'' people. These people work in parallel to complete the shuttle. A parallel computer is quite similar, we might use 100,000 digital computers working together to simulate airflow over a new shuttle design. Key in NASA's shuttle project is the management structure. This becomes, for the analogy, the issue of computer hardware and software architecture; a key research area in computer science.
We can view the brain as a parallel computer with some 
neurons working together to solve information processing and
decision-making problems. The neurons are analogous to the node shown
in Figure 1(a); nature links neurons by axons and
dendrites, not wires and printed circuit board traces used by nCUBE.
However, the basic design---interconnected elements communicating by
message passing---is the same and further both nature's and digital
parallel computer use the same mechanism of data parallelism to solve
problems concurrently.