Ian Glendinning

The MPI Message Passing Interface Standard

The diverse message passing interfaces provided on parallel and distributed computing systems have caused difficulty in movement of application software from one system to another and have inhibited the commercial development of tools and libraries for these systems. The Message Passing Interface (MPI) Forum has developed a de facto interface standard which was finalised in Ql of 1994. Major parallel system vendors and software developers were involved in the definition process, and the first implementations of MPI are already appearing. This article presents an overview of the MPI initiative and the standard interface, in particular those aspects which merge demonstrated research with common practice.

Practical Parallelism using Transputer Arrays

This paper explores methods for extracting parallelism from a wide variety of numerical applications. We investigate communications overheads and load-balancing for networks of transputers. After a discussion of some practical strategies for constructing occam programs, two case studies are analysed in detail.

Parallelization of the QC-lib Quantum Computer Simulator Library

We report on work to parallelize QC-lib, a C++ library for the simulation of quantum computers at an abstract functional level. After a brief introduction to quantum computing, we give an outline of QC-lib, then describe its parallelization using MPI, and present performance measurements made on a Beowulf cluster. Using more processors allowed larger problems to be solved, and reasonable speedups were obtained for the Hadamard transform and Grover’s quantum search algorithm.

THE JADE MUON DETECTOR

Abstract The JADE muon detector consists of 618 planar drift chambers interspersed between layers of hadron absorber. This paper gives a detailed description of the construction and operation of the detector as a whole and discusses the properties of the drift chambers. The muon detector has been operating successfully at PETRA for five years.

MUON IDENTIFICATION IN JADE

Abstract The method of identification of high energy muons in the JADE detector is described in detail. The performance of the procedure is discussed in detail for the case of prompt muon identification in multihadronic final states.

Transputer arrays as FORTRAN farms for particle physics

Abstract The INMOS transputer, the T414, and the newly-announced floating-point version, the T800, capable of speeds in excess of one Mflop, are powerful and versatile microprocessors which are natural building blocks for multiprocessor arrays. This paper details our experience using transputer arrays in a FORTRAN “farm” mode, in which each transputer runs the same program on different data, and passes the results to a host microcomputer. The transputer hardware used consists of a Meiko Computing Surface machine. Benchmark timings for the T414 are given for the LUND Monte Carlo generator FORTRAN program used to model QCD jet events. Preliminary results obtained using engineering samples of the T800 are also given.

Parallelisation of a Satellite Signal Processing Code - Strategies and Tools

This paper presents strategies and tools that have been used in work to parallelise a satellite signal processing code. The Magellan mission to map the surface of Venus using synthetic aperture radar (SAR) is briefly described, the Magellan SAR processor code is outlined, and a parallelisation strategy is presented. The codes large size and limited documentation made the use of program analysis tools essential to implement this strategy. Three tools, FORESYS, IDA and FORGExplorer, are compared, and the use of FORGExplorer to perform code analysis is described in detail. The techniques presented are of general applicability to the parallelisation of codes in other application areas.