The duality of memory and communication in the implementation of a multiprocessor operating system
SOSP '87 Proceedings of the eleventh ACM Symposium on Operating systems principles
Applications experience with Linda
PPEALS '88 Proceedings of the ACM/SIGPLAN conference on Parallel programming: experience with applications, languages and systems
The design and analysis of parallel algorithms
The design and analysis of parallel algorithms
Programming languages for distributed computing systems
ACM Computing Surveys (CSUR)
Simple but effective techniques for NUMA memory management
SOSP '89 Proceedings of the twelfth ACM symposium on Operating systems principles
SOSP '89 Proceedings of the twelfth ACM symposium on Operating systems principles
Mirage: a coherent distributed shared memory design
SOSP '89 Proceedings of the twelfth ACM symposium on Operating systems principles
Memory coherence in shared virtual memory systems
ACM Transactions on Computer Systems (TOCS)
Computer
ACM Transactions on Computer Systems (TOCS)
Concepts and Notations for Concurrent Programming
ACM Computing Surveys (CSUR)
Distributed data structures in Linda
POPL '86 Proceedings of the 13th ACM SIGACT-SIGPLAN symposium on Principles of programming languages
Experimental Comparison of Memory Management Policies for NUMA Multiprocessors
Experimental Comparison of Memory Management Policies for NUMA Multiprocessors
Shared virtual memory on loosely coupled multiprocessors
Shared virtual memory on loosely coupled multiprocessors
| Hi-index | 0.00 |
In this paper we propose a general framework of distributed computing in a NUMP environment. A NUMP (Non-Uniform Message-Passing) distributed environment is a computing environment in which both tightly-coupled, shared-memory systems and loosely-coupled, distributed computers are co-existent, and the time for message-passing is non-uniform between various systems. An environment with such hybrid components has become a common phenomenon in most academic and industrial communities. The issue of how to develop and execute programs that take advantage of all the resources available in a hybrid environment is an open issue in the research that carries both theoretic and pragmatic significance. The main topics involved include the development of new computational models for NUMP environment, operating system and language support for NUMP computing, methodologies in NUMP programming, principles of algorithm design for NUMP system, and so on. We deem that the research in NUMP computing will contribute in providing a unified framework for parallel and distributed computing, and in accomplishing better utilization of computing resources at large scale.