Locality in parallel computation. Master's thesis

Technical Report · Fri Sep 01 04:00:00 EDT 1989

OSTI ID:6999295

Maggs, B M

This thesis explores strategies for exploiting locality in three major areas of parallel computation: packet routing, graph algorithms, and network emulations. Chapter 1 describes a network-independent approach to the packet-routing problem. Our strategy is to partition the problem into two stages: a path-selection stage and a scheduling stage. Chapter 2 introduces a model for parallel computation, called the distribution random-access machine (DRAM), in which the communication requirements of parallel computer in which memory accesses are evaluated. A DRAM is an abstraction of a parallel computer in which memory accesses are implemented by routing messages through a communication network. Chapter 3 examines the problem of how efficiently a host network can emulate a guest network. The goal is to emulate TG steps of an NG-node guest network on an NH node host network. Chapter 4 presents an algorithm for finding a minimum-cost spanning tree of an N-node graph on an N x N mesh-connected computer. The algorithm has the same O(N) running time as the previous algorithms, but it is much simpler.

• Research Organization: Massachusetts Inst. of Tech., Cambridge, MA (USA). Lab. for Computer Science
• OSTI ID: 6999295
• Report Number(s): AD-A-218733/4/XAB; MIT/LCS/TR--469
• Country of Publication: United States
• Language: English

Related Subjects

99 GENERAL AND MISCELLANEOUS
990200* -- Mathematics & Computers
ALGORITHMS
COMMUNICATIONS
COMPUTERS
DATA PROCESSING
DISTRIBUTED DATA PROCESSING
DISTRIBUTION
GRAPHS
LOCALITY
MATHEMATICAL LOGIC
MEMORY DEVICES
PARALLEL PROCESSING
PROCESSING
PROGRAMMING
ROUTING
SPECIFICATIONS