Abstract
Even though distributed memory machines are the prevailing supercomputer architecture due to their superior scalability, programming for this class of parallel computers is still cumbersome and inefficient compared to sequential programming. Also resource utilization is often poor in comparison with sequential machines. This results from a lack of support by system software, i.e. compilers, operating systems, and programming environments. Especially operating systems are traditionally responsible to take some burden from the programmer by automatically managing resources and providing a logical, comfortable interface for resource usage. To achieve efficient operation, we need an operating system that takes care of the peculiarities of multicomputer systems. Since the efficiency of interprocessor communication is a key factor for overall performance, the operating system should add as little overhead as possible to communication. To meet this goal, the common approach is to give up the concept of a resident operating system and to provide the necessary functionality by a run-time library linked to the parallel application program. This certainly has the desired effect but is accompanied by other disadvantages:
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: 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.