Application-centric parallel multimedia software

Parallel and distributed high-performance computing will be generally available both implicitly and explicitly for digital multimedia applications. This is not a conjecture; it's a reflection of current activity in engineering systems for visual- and data-intensive multimedia applications. As a result, parallel high-performance computing must mature very quickly and come into the mainstream of software development. For this to happen, research in areas of parallel computing software such as languages, system designs, compilers, and tools construction must break out of the mold and reestablish the objective of high-performance parallel computing for the masses. The focus needs to be on what parallel computing can do for the applications instead of what the applications can do with parallel computing. The latter principle has governed most parallel computing efforts to date. Digital multimedia computing is expected to achieve consumer status very rapidly. This dictates that the development of high-performance parallel computing multimedia applications must be as easy and integrated to everyday computing environments as is the development of Java applets. In other words, one needs a user-oriented and application-centric view of parallel computing software, instead of the system-oriented and number-centric focus that has dominated past and current development