Concurrency for algorithms shall exploit the many core architecture of current CPUs or even explore many machines (nodes) available. Concurrency can be interpre

Concurrency for algorithms shall exploit the many core architecture of current CPUs or even explore many machines (nodes) available. Concurrency can be interpreted as a way of optimization, and one shall start with optimization only at the bottle necks (zen wisedom). That's why parallelism is often applied at compute intense parts of the code (there is even a science about that: High performance computing). Map-Reduce is one of the classic and most widespread approaches.