Abstract
Today’s sequential processors are not designed in a vacuum. Instead, they are built to fulfill the needs of extensive sets of benchmark programs and at the same time take into account the capabilities of compilers. A comparable level of maturity has not been reached in the design of parallel computers. The semantic gap between parallel hardware and high-level, parallel languages is substantial at present, and far too large to be bridged effectively by a compiler. The result is that programmers must code at a low, machine-oriented level and that parallel programs are largely non-portable. This poor state of affairs is not surprising, given that many of the variables involved in parallel system design are unknown and in a state of flux. These variables include the capabilities that parallel machines can offer, the translation and optimization techniques of compilers for parallel machines, and the appropriate high-level constructs in parallel programming languages. In addition, parallel system architecture allows many more degrees of freedom than sequential systems. In the long run, however, the practice of rewriting parallel programs for every new machine architecture is economically intolerable.
Sign-in/Register to access full text options 
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 callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
