Thomas Bemmerl

An Integrated Environment for Programming Distributed Memory Multiprocessors

A breakthrough of distributed memory architectures as general purpose programmable computers depends strongly on the ease of programming such machines. TOPSYS (TOols for Parallel SYStems) is an integrated environment for programming distributed memory multiprocessors. Its concepts and application on an iPSC/2 multiprocessor are presented. It is based on a simple message passing process model implemented in the operating system kernel MMK with a global object space. A smooth transition of parallelization methodology from manual to interactive (semiautomated) and finally to fully automated is provided with this environment. It uses several distributed monitoring techniques and offers various development tools ranging from specification to performance analysis. Fully transparent dynamic load balancing is being implemented.

The TOPSYS Architecture

A survey on the TOPSYS (TOols for Parallel SYStems) project at the Department of Computer Science at Technical University of Munich is presented. Within this project, an integrated tool environment for increasing programmer productivity when using and programming parallel computers is developed. The TOPSYS tool environment offers tools for specification, mapping, debugging, testing, performance analysis, graphical program animation and dynamic loadbalancing of parallel programs. In addition to these tools a distributed operating system kernel and a synthetic workload generator has been developed. Apart from the integrated hierarchical architecture, the major features of the TOPSYS environment are the support of different monitoring techniques, easy adaptability, high portability and a common graphic user interface for all tools. After the description of the project goals, the major design concepts and the state of the project we describe a first application of the TOPSYS tools.

MMK—a distributed operating system kernel with integrated dynamic loadbalancing

The paper presents an operating system kernel for highly parallel supercomputers, which was implemented on an iPSC/2 Hypercube with 32 processors. The kernel offers a process model, which is well suited for most partitioning strategies of parallel algorithms. The base for the efficiency of this object oriented, global, and dynamic programming concept are advances in communication network technologies (virtual fully connection) of some new parallel supercomputers. After presenting the functionality and the implementation of MMK (Multiprocessor Multitasking Kernel), the paper reports on an improved programming methodology based on a combination of data and task partitioning which leads to efficient computations on virtual fully connected highly parallel machines. MMK is an integral part of the TOPSYS-project (TOols for Parallel SYStems) and all tools support the MMK programming model.

Realtime high level debugging in host/target environments

Abstract Programs for embedded systems with 16- or 32-bit microprocessors, often realtime systems, are usually written in a high level programming language and are based on a realtime multitasking kernel. But, available microprocessor development systems do not sufficiently support realtime debugging of such software at a high level. Only tools for realtime debugging at a level which lies between assembler and high level languages, are available. This paper presents new concepts for realtime high level debugging tools in host/target environments. In addition a prototype of a realtime high level debugging system is introduced, which allows debugging of programs written in a high level language (C or Pascal) and based on an object oriented realtime multitasking kernel in the target microcomputer.

Menu and graphic driven human interfaces for high level debuggers

Abstract This paper deals with the design of adequate human interfaces for debuggers. First we will point out the designing problems and describe some solutions we found in our studies. Then we will illustrate our approaches to these problems as we designed our graphical human interfaces. At the end of the paper we will give some details about implementing two interfaces. These interfaces are based on different graphic subsystems. MacBug is realized on top of the graphic primitives of the Macintosh and GUI is run on the VAXstation II and based on X-Windows. Both human interfaces are frontends for the concurrent high level debugger REALBUG.

Experiences in Parallelizing an Existing CFD Algorithm

The numerical simulation of a 2- or 3-dimensional motion of viscous multiphase flow can be described by a system of partial differential equations. Because this simulation poses extremely high demands on computing capacity it is very well suited for calculation by a multiprocessor. This paper describes the implementation of FLUBOX, a two-dimensional fluid dynamics code developed at the Gesellschaft fur Reaktorsicherheit on an iPSC/2 distributed memory multiprocessor. The main criterion for evaluation was to get a good speedup by changing only small parts of the program. The program has been analyzed using the performance analyzer PATOP developed in the TOPSYS project. In this way, we were able to study the dynamic behaviour of the algorithm.

Quellbezogenes Debugging von Multimikroprozessoren

Der Artikel beschreibt Anforderungen, Eigenschaften, Entwurfs- und Implementierungskonzepte von Werkzeugen zur Fehlersuche (Debugging) in Multimikroprozessoren und deren Programmen. Besonderes Gewicht wird dabei auf die Unterstutzung hoherer Abstraktionskonzepte (Hochsprachen, konkurrente Spracherweiterungen) gelegt. Ein weiterer Schwerpunkt ist die Integration verschiedener Instrumentierungstechniken (Hardware-, Software- und Hybrid-Instrumentierung) in ein Debugging System. Die Entwurfskonzepte und Implementierungsdetails von Debugging Systemen fur Multimikroprozessoren werden exemplarisch anhand des an der TU Munchen entwickelten Debugging Systems REALBUG aufgezeigt. Neben der Beschreibung der von REALBUG angebotenen Funktionalitaten und Abstraktionen werden insbesondere Konzepte zur Erlangung hoherer Portabilitat, Erweiterbarkeit und Adaptierbarkeit von quellbezogenen Uberwachungswerkzeugen vorgestellt.

Programming tools for distributed multiprocessor computing environments

Abstract With the increasing availability of multiprocessor platforms based on different types of architectures (shared memory, distributed memory, network based), users will be increasingly faced with heterogeneous and distributed multiprocessor computing facilities. Programming environment concepts have to be found which enable the user to program and use these heterogeneous computing resources, consisting of different multiprocessor architectures, in a transparent manner. Topics which have to be addressed and tools which have to be developed are parallel programming models, monitoring systems, debuggers, performance analyzers, animation systems and dynamic load-balancing schemes. The paper describes the integrated tool environment TOPSYS (TOols for Parallel SYStems), available for the iPSC/2 and iPSC/860 hypercube, as well as its adaptation to a network of UNIX-based workstations.

Microprocessor architectures: a basis for real-time systems

Four years after the “birth” of the microprocessor in 1971, Fortune Magazine wrote the following about the effects of this technology on human life:

Integrierter Einsatz von Parallelrechnern mit gemeinsamem und verteiltem Speicher - Alliant FX/2800, iPSC/2, iPSC/860*

Parallelrechner mit unterschiedlichster Architektur sind heute auf dem Markt erhaltlich. Im Laufe der Zeit haben sich zwei Klassen von Multiprozessoren herauskristallisiert, welche man in naher Zukunft in beinahe allen Rechenzentren vorfinden wird — Multiprozessoren mit gemeinsamem und verteiltem Speicher. Es stellt sich damit die Frage, wie diese Systeme in existierende netzwerkbasierte Rechnerumgebungen integriert und kooperierend mit konventionellen (sequentiellen) Rechensystemen eingesetzt werden konnen. Der vorliegende Artikel beschreibt am Beispiel der Systeme Alliant FX/2800 und iPSC die Architektur, Betriebssysteme und Programmierung beider Klassen von Multiprozessoren. Am Beispiel einer an der Technischen’ Universitat Munchen installierten Konfiguration beider Systemklassen wird eine integrierte und kooperierende Nutzung und Programmierung mit Hilfe der Werkzeugumgebung TOPSYS (TOols for Parallel SYStems) vorgeschlagen.