Portability, predictability and performance for parallel computing: BSP in practice

We report on practical experience using the Oxford BSP Library to parallelize a large electromagnetic code, the British Aerospace finite-difference time-domain code EMMA T:FD3D. The Oxford BS Library is one of the first realizations of the Bulk Synchronous Parallel computational model to be targeted at numerically intensive scientific (typically Fortran) computing. The BAe EMMA code is one of the first large-scale applications to be parallelized using this library, and it is an important demonstration of the cost effectiveness of the BSP approach. We illustrate how BSP cost-modelling techniques can be used to predict and optimize performance for single-source programs across different parallel platforms. We provide predicted and observed performance figures for an industrial-strength, single-source parallel code for a variety of real parallel architectures: shared memory multiprocessors, workstation clusters and massively parallel platforms.

[Book review] Rosalind W. Picard, Affective Computing

Review of: Rosalind W. Picard, Affective Computing

Multilevel mesh partitioning for optimizing domain shape

Multilevel algorithms are a successful class of optimization techniques that address the mesh partitioning problem for mapping meshes onto parallel computers. They usually combine a graph contraction algorithm together with a local optimization method that refines the partition at each graph level. To date, these algorithms have been used almost exclusively to minimize the cut-edge weight in the graph with the aim of minimizing the parallel communication overhead. However, it has been shown that for certain classes of problems, the convergence of the underlying solution algorithm is strongly influenced by the shape or aspect ratio of the subdomains. Therefore, in this paper, the authors modify the multilevel algorithms to optimize a cost function based on the aspect ratio. Several variants of the algorithms are tested and shown to provide excellent results.

Bulk synchronous parallelization of industrial electromagentic software

The parallelization of existing/industrial electromagnetic software using the bulk synchronous parallel (BSP) computation model is presented. The software employs the finite element method with a preconditioned conjugate gradient-type solution for the resulting linear systems of equations. A geometric mesh-partitioning approach is applied within the BSP framework for the assembly and solution phases of the finite element computation. This is combined with a nongeometric, data-driven parallel quadrature procedure for the evaluation of right-hand-side terms in applications involving coil fields. A similar parallel decomposition is applied to the parallel calculation of electron beam trajectories required for the design of tube devices. The BSP parallelization approach adopted is fully portable, conceptually simple, and cost-effective, and it can be applied to a wide range of finite element applications not necessarily related to electromagnetics.

Shape-optimized mesh partitioning and load balancing for parallel adaptive FEM

We present a dynamic distributed load balancing algorithm for parallel, adaptive Finite Element simulations in which we use preconditioned Conjugate Gradient solvers based on domain-decomposition. The load balancing is designed to maintain good partition aspect ratio and we show that cut size is not always the appropriate measure in load balancing. Furthermore, we attempt to answer the question why the aspect ratio of partitions plays an important role for certain solvers. We define and rate different kinds of aspect ratio and present a new center-based partitioning method of calculating the initial distribution which implicitly optimizes this measure. During the adaptive simulation, the load balancer calculates a balancing flow using different versions of the diffusion algorithm and a variant of breadth first search. Elements to be migrated are chosen according to a cost function aiming at the optimization of subdomain shapes. Experimental results for Bramble's preconditioner and comparisons to state-of-the-art load balancers show the benefits of the construction.

Using mathematical models to predict the development of aircraft cabin fires

Computer based mathematical models describing aircraft fire have a role to play in the design and development of safer aircraft, in the implementation of safer and more rigorous certification criteria and in post mortuum accident investigation. As the cost involved in performing large-scale fire experiments for the next generation 'Ultra High Capacity Aircraft' (UHCA) are expected to be prohibitively high, the development and use of these modelling tools may become essential if these aircraft are to prove a safe and viable reality. By describing the present capabilities and limitations of aircraft fire models, this paper will examine the future development of these models in the areas of large scale applications through parallel computing, combustion modelling and extinguishment modelling.

Dynamic multi-partitioning for parallel finite element applications

The central product of the DRAMA (Dynamic Re-Allocation of Meshes for parallel Finite Element Applications) project is a library comprising a variety of tools for dynamic re-partitioning of unstructured Finite Element (FE) applications. The input to the DRAMA library is the computational mesh, and corresponding costs, partitioned into sub-domains. The core library functions then perform a parallel computation of a mesh re-allocation that will re-balance the costs based on the DRAMA cost model. We discuss the basic features of this cost model, which allows a general approach to load identification, modelling and imbalance minimisation. Results from crash simulations are presented which show the necessity for multi-phase/multi-constraint partitioning components

Computing regular equivalence: practical and theoretical issues

Social network analysts have tried to capture the idea of social role explicitly by proposing a framework that precisely gives conditions under which group actors are playing equivalent roles. They term these methods positional analysis techniques. The most general definition is regular equivalence which captures the idea that equivalent actors are related in a similar way to equivalent alters. Regular equivalence gives rise to a whole class of partitions on a network. Given a network we have two different computational problems. The first is how to find a particular regular equivalence. An algorithm exists to find the largest regular partition but there are not efficient algorithms to test whether there is a regular k-partition. That is a partition in k groups that is regular. In addition, when dealing with real data, it is unlikely that any regular partitions exist. To overcome this problem relaxations of regular equivalence have been proposed along with optimisation techniques to find nearly regular partitions. In this paper we review the algorithms that have developed to find particular regular equivalences and look at some of the recent theoretical results which give an insight into the complexity of finding regular partitions.

The same but different – how to introduce variation within computing assessment tasks

In this article, suggestions are made for introducing an individual element into formative assessment of the ability to use computer software for statistics.

Computer Museum, School of Computing and Mathematical Sciences, University of Greenwich

Computer equipment, once viewed as leading edge, is quickly condemned as obsolete and banished to basement store rooms or rubbish bins. The magpie instincts of some of the academics and technicians at the University of Greenwich, London, preserved some such relics in cluttered offices and garages to the dismay of colleagues and partners. When the University moved into its new campus in the historic buildings of the Old Royal Naval College in the center of Greenwich, corridor space in King William Court provided an opportunity to display some of this equipment so that students could see these objects and gain a more vivid appreciation of their subject's history.

Self-adaptability and man-in-the-loop: a dilemma in autonomic computing systems

This paper addresses some controversial issues relating to two main questions. Firstly, we discuss 'man-in-the loop' issues in SAACS. Some people advocate this must always be so that man's decisions can override autonomic components. In this case, the system has two subsystems - man and machine. Can we, however, have a fully autonomic machine - with no man in sight; even for short periods of time? What kinds of systems require man to always be in the loop? What is the optimum balance in self-to-human control? How do we determine the optimum? How far can we go in describing self-behaviour? How does a SAACS system handle unexpected behaviour? Secondly, what are the challenges/obstacles in testing SAACS in the context of self/human dilemma? Are there any lesson to be learned from other programmes e.g. Star-wars, aviation and space explorations? What role human factors and behavioural models play whilst in interacting with SAACS?.

Parallel implementation of the recurrence method for computing the power-spectral density of thin avalanche photodiodes

A simulation program has been developed to calculate the power-spectral density of thin avalanche photodiodes, which are used in optical networks. The program extends the time-domain analysis of the dead-space multiplication model to compute the autocorrelation function of the APD impulse response. However, the computation requires a large amount of memory space and is very time consuming. We describe our experiences in parallelizing the code using both MPI and OpenMP. Several array partitioning schemes and scheduling policies are implemented and tested Our results show that the OpenMP code is scalable up to 64 processors on an SGI Origin 2000 machine and has small average errors.

A policy-definition language and prototype implementation library for policy-based autonomic systems

This paper presents work towards generic policy toolkit support for autonomic computing systems in which the policies themselves can be adapted dynamically and automatically. The work is motivated by three needs: the need for longer-term policy-based adaptation where the policy itself is dynamically adapted to continually maintain or improve its effectiveness despite changing environmental conditions; the need to enable non autonomics-expert practitioners to embed self-managing behaviours with low cost and risk; and the need for adaptive policy mechanisms that are easy to deploy into legacy code. A policy definition language is presented; designed to permit powerful expression of self-managing behaviours. The language is very flexible through the use of simple yet expressive syntax and semantics, and facilitates a very diverse policy behaviour space through both hierarchical and recursive uses of language elements. A prototype library implementation of the policy support mechanisms is described. The library reads and writes policies in well-formed XML script. The implementation extends the state of the art in policy-based autonomics through innovations which include support for multiple policy versions of a given policy type, multiple configuration templates, and meta-policies to dynamically select between policy instances and templates. Most significantly, the scheme supports hot-swapping between policy instances. To illustrate the feasibility and generalised applicability of these tools, two dissimilar example deployment scenarios are examined. The first is taken from an exploratory implementation of self-managing parallel processing, and is used to demonstrate the simple and efficient use of the tools. The second example demonstrates more-advanced functionality, in the context of an envisioned multi-policy stock trading scheme which is sensitive to environmental volatility

Exploring adaptation & self-adaptation in autonomic computing systems

This panel paper sets out to discuss what self-adaptation means, and to explore the extent to which current autonomic systems exhibit truly self-adaptive behaviour. Many of the currently cited examples are clearly adaptive, but debate remains as to what extent they are simply following prescribed adaptation rules within preset bounds, and to what extent they have the ability to truly learn new behaviour. Is there a standard test that can be applied to differentiate? Is adaptive behaviour sufficient anyway? Other autonomic computing issues are also discussed.