Abstract
Parallel architectures will become the standard form of computers only if parallel programming can be made as easy as sequential programming. With the existing programming models we still are far from that goal. Therefore, intensive research and development is needed in the areas of high-level programming models and programming tools, parallelizing compilers, and architectures of node hardware and operating systems that support the high-level programming models and parallelizing compilers. In the paper the state-of-the-art of node hardware and operating system architecture is depicted. The suitability of the existing programming models is discussed, and a look is taken at their supporting architectures, as well as the existing parallel programming tools. This demonstrates the rationale for raising the level of abstraction of parallel programming paradigms. Examples for such high-level programming models are given. The state of the art of parallelizing compilers is outlined, and it is discussed how the task of such compilers can be supported by new architectural concepts.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
