Memory coherence in shared virtual memory systems
Kai Li, Paul Hudak
Abstract
Open-access reader
Kai Li, Paul Hudak
Abstract
Open-access reader
The memory coherence problem in designing and implementing a shared virtual memory on loosely coupled multiprocessors is studied in depth. Two classes of algorithms, centralized and distributed, for solving the problem are presented. A prototype shared virtual memory on an Apollo ring based on these algorithms has been implemented. Both theoretical and practical results show that the memory coherence problem can indeed be solved efficiently on a loosely coupled multiprocessor.
OpenAlex reports 1171 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The memory coherence problem in designing and implementing a shared virtual memory on loosely coupled multiprocessors is studied in depth. Two classes of algorithms, centralized and distributed, for solving the problem are presented. A prototype shared virtual memory on an Apollo ring based on these algorithms has been implemented. Both theoretical and practical results show that the memory coherence problem can indeed be solved efficiently on a loosely coupled multiprocessor.
Key concepts: Computer science, Distributed shared memory, Shared memory, Distributed memory, Data diffusion machine, Parallel computing, Multiprocessing, Cache coherence