Embarassingly parallel

In the jargon of parallel computing, an embarassingly parallel workload is one for which no particular effort is needed to segment the problem into parallel tasks, and there is very little dependency between those parallel tasks. Examples of embarassingly parallel problems include non-real-time computer graphics rendering, brute-force searches in cryptography, and tasks like SETI@home.

Embarassingly parallel problems are ideally suited to distributed computing over the Internet, and are also easy to perform on server farms which do not have any of the special infrastructure used in a true supercomputer cluster.