Competitive snoopy caching for large-scale multiprocessors
作者： Mark S. Manasse
DOI:
关键词: Block (data storage) 、 Snoopy cache 、 Scale (descriptive set theory) 、 Default gateway 、 Hypercube 、 Computer network 、 Computer science 、 Line (text file)
摘要: A system and method of satisfying read write requests is used in a having plurality cache-equipped processors coupled into hypercube structure via buses, where each processor simultaneously to other on buses gateway means. Read for line data from any the are satisfied by forwarding an update or invalidate request given block containing requested. This forwarded all which present. The present invention provides responding request, resultant data, one stored, responds forward along exactly branch.
lens.org 参考文章(7)
Carl Feyman, Tamiko Thiel, Brewster Kahle, Richard Clayton, W. D. Hillis, Method and apparatus for interconnecting processors in a hyper-dimensional array ,(1986)
Daniel Mark Nosenchuck, Michael Geist Littman, Multinode reconfigurable pipeline computer ,(1987)
Renben Shu, David H. C. Du, Improved hypercube topology for multiprocessor computer systems ,(1991)
W. Daniel Hillis, Carl R. Feynman, David P. Christman, Cliff A. Lasser, Brewster L. Kahle, Alan Bawden, David Chapman, Thomas F. Knight, Parallel processor/memory circuit ,(1983)
Francis E. Rix, David Hayes, Multi-dimensional-access memory system with combined data rotation and multiplexing ,(1983)
Don Lieberman, Jesus O. Tuazon, John C. Peterson, Moshe Pniel, Concurrent hypercube system with improved message passing ,(1986)
Erik P. DeBenedictis, Parallel processing network and method ,(1986)