A VLSI Architecture for Concurrent Data Structures (The Springer International Series in Engineering and Computer Science)

Dally, J. W.

ISBN 10: 0898382351 / ISBN 13: 9780898382358

Editore: Springer, 1987

Riguardo questo articolo

Very Good.

Codice articolo ERICA75808983823515

Dati bibliografici

• Titolo: A VLSI Architecture for Concurrent Data Structures (The Springer International Series in Engineering and Computer Science)
• Editore: Springer
• Data di pubblicazione: 1987
• Lingua: Inglese
• ISBN 10: 0898382351
• ISBN 13: 9780898382358
• Rilegatura: Hardcover
• Condizione: Very Good
• Tipo di libro: book

Descrizione articolo

Riassunto

Concurrent data structures simplify the development of concurrent programs by encapsulating commonly used mechanisms for synchronization and commu­ nication into data structures. This thesis develops a notation for describing concurrent data structures, presents examples of concurrent data structures, and describes an architecture to support concurrent data structures. Concurrent Smalltalk (CST), a derivative of Smalltalk-80 with extensions for concurrency, is developed to describe concurrent data structures. CST allows the programmer to specify objects that are distributed over the nodes of a concurrent computer. These distributed objects have many constituent objects and thus can process many messages simultaneously. They are the foundation upon which concurrent data structures are built. The balanced cube is a concurrent data structure for ordered sets. The set is distributed by a balanced recursive partition that maps to the subcubes of a binary 7lrcube using a Gray code. A search algorithm, VW search, based on the distance properties of the Gray code, searches a balanced cube in O(log N) time. Because it does not have the root bottleneck that limits all tree-based data structures to 0(1) concurrency, the balanced cube achieves 0C.:N) con­ currency. Considering graphs as concurrent data structures, graph algorithms are pre­ sented for the shortest path problem, the max-flow problem, and graph parti­ tioning. These algorithms introduce new synchronization techniques to achieve better performance than existing algorithms.

Contenuti

1 Introduction.- 1.1 Original Results.- 1.2 Motivation.- 1.3 Background.- 1.4 Concurrent Computers.- 1.4.1 Sequential Computers.- 1.4.2 Shared-Memory Concurrent Computers.- 1.4.3 Message-Passing Concurrent Computers.- 1.5 Summary.- 2 Concurrent Smalltalk.- 2.1 Object-Oriented Programming.- 2.2 Distributed Objects.- 2.3 Concurrency.- 2.4 Locks.- 2.5 Blocks.- 2.6 Performance Metrics.- 2.7 Summary.- 3 The Balanced Cube.- 3.1 Data Structure.- 3.1.1 The Ordered Set.- 3.1.2 The Binary nCube.- 3.1.3 The Gray Code.- 3.1.4 The Balanced Cube.- 3.2 Search.- 3.2.1 Distance Properties of the Gray Code.- 3.2.2 VW Search.- 3.3 Insert.- 3.4 Delete.- 3.5 Balance.- 3.6 Extension to B-Cubes.- 3.7 Experimental Results.- 3.8 Applications.- 3.9 Summary.- 4 Graph Algorithms.- 4.1 Nomenclature.- 4.2 Shortest Path Problems.- 4.2.1 Single Point Shortest Path.- 4.2.2 Multiple Point Shortest Path.- 4.2.3 All Points Shortest Path.- 4.3 The Max-Flow Problem.- 4.3.1 Constructing a Layered Graph.- 4.3.2 The CAD Algorithm.- 4.3.3 The CVF Algorithm.- 4.3.4 Distributed Vertices.- 4.3.5 Experimental Results.- 4.4 Graph Partitioning.- 4.4.1 Why Concurrency is Hard.- 4.4.2 Gain.- 4.4.3 Coordinating Simultaneous Moves.- 4.4.4 Balance.- 4.4.5 Allowing Negative Moves.- 4.4.6 Performance.- 4.4.7 Experimental Results.- 4.5 Summary.- 5 Architecture.- 5.1 Characteristics of Concurrent Algorithms.- 5.2 Technology.- 5.2.1 Wiring Density.- 5.2.2 Switching Dynamics.- 5.2.3 Energetics.- 5.3 Concurrent Computer Interconnection Networks.- 5.3.1 Network Topology.- 5.3.2 Deadlock-Free Routing.- 5.3.3 The Torus Routing Chip.- 5.4 A Message-Driven Processor.- 5.4.1 Message Reception.- 5.4.2 Method Lookup.- 5.4.3 Execution.- 5.5 Object Experts.- 5.6 Summary.- 6 Conclusion.- A Summary of Concurrent Smalltalk.- B Unordered Sets.- B.1 Dictionaries.- B.2 Union-Find Sets.- C On-Chip Wire Delay.