Axel Lehmann

Performance evaluation and prediction of storage hierarchies

This paper proposes a modelling methodology combining simulation and analysis for computer performance evaluation and prediction. The methodology is based on a special workload model that is suitable for the generation and description of dynamic program behaviour. A description of this workload model is given in section 2. The applicability of this concept with respect to the design of new storage systems, as well as the improvement or comparison of existing systems, will be described by investigation of the efficiency of small cache memories in section 3.

Knowledge-Based Modelling and Simulation: Restrictions, Alternatives and Applications

This paper gives a survey of different categories of knowledge-based systems or expert systems, respectively, for supporting modelling and simulation of system dynamics in discrete systems. At first, actual requirements concerning support of users in model construction and goal-directed, efficient application of models are summarized. Regarding similarities existing between expert systems and simulation models with respect to their intended purpose, principal approaches for the combined and adjusted application of expert system and simulation techniques are discussed in this paper. As a result, knowledge-based modelling and simulation environments are classified regarding their taxonomy and their intended applicability. Some examples for practical applications of expert systems as classification and advisory systems for user supporting tools in the modelling process are finally presented.

Leistungsanalyse mit INT3: Einer Interaktiven, Intelligenten und Integrierten PC-Modellierungsumgebung

Dieser Beitrag beschreibt ein Konzept und die Realisierung eines Prototyps zur interaktiven, wissensbasierten Unterstutzung von Anwendern in den verschiedenen Phasen eines Modellierungsprozesses. Das vorliegende Konzept geht von der Annahme aus, das in zunehmendem Mase Systemspezialisten aus verschiedensten Anwendungsbereichen detaillierte Systemanalysen als Grundlage fur Entscheidungen benotigen, die den praktischen Einsatz dieser Systeme, deren Entwurf, Auswahl und Modifikation betreffen. Es wird davon ausgegangen, das die meisten dieser Systemspezialisten nur uber Grundkenntnisse im Bereich der Systemanalyse verfugen, einen mangelnden Uberblick uber Modellierungsmethoden und -Werkzeuge besitzen, aber andererseits mit zunehmender Verbreitung von PC’s einen direkten Zugang zu leistungsfahigen Verarbeitungssystemen haben.

Practical experiences with vertical migration

Abstract The performance of microprogrammable computer system architectures can be improved considerably by microcoding frequently executed and CPU intensive software functions. This paper reports on quantitative performance enhancements obtained by practical application of vertical migration techniques in a hierarchically layered software/firmware/hardware system Burroughs B 1726. Various representative, CPU intensive and runtime intensive software primitives have been selected by measurements — operating system primitives as well as application software primitives —, have been microcoded and migrated to the firmware layer. Quantitative analyses summarized in this presentation demonstrate the range of performance enhancements which can be expected by the application of this technique.

Effective values: an approach for characterizing dependability parameters

The authors discuss a novel direction for evaluating system dependability. The research efforts outlined are part of a special focus concerning the dependability of complex distributed systems at the Department of Computer Engineering of the Federal Armed Forces University in Munich, Germany. Our approach emphasizes the behavior of components after they have been embedded in a system. We use the notion of effective values for the analysis and estimation of key dependability parameters. The concept is targeted to unify performance characterization with the dependability. Besides analysis, we report on some simulation results, which were in favor of our assumptions.

Computer Aided Modelling on PC with INT3 (Interactive, Intelligent and Integrated Modelling Environment)

This paper presents INT3: an interactive, intelligent and integrated tool supporting performance modelling in a standard IBM-PC/AT environment. INT3 provides the user with a graphical interface and together with integrated expert-system components, assists him interactively, to pass through all phases of a modelling process:

Leistungsfähigkeit von Mehrprozessorsystemen mit iAPX 432-Prozessoren, Kreuzschienenverteiler und Pufferspeichern

In dem vorliegenden Beitrag wird die Leistungsfahigkeit von Mehrprozessorsystemen mit Prozessoren des Typs INTEL iAPX 432 quantitativ bewertet. Die Leistungsbewertungen basieren auf Simulationen der Programm- und Organisationsablaufe, die bei objektori-entierter und nebenlaufiger Verarbeitung zwischen Verarbeitungseinheiten und Arbeits-speicher auf der Maschinenbefehlsebene zu beobachten sind.

Leistungsanalyse von Speicherhierarchien durch hybride und hierarchische Modellierung

Rasche technologische Fortschritte, kurzere Innovationszyklen und steigende Systemkomplexitat bestimmen die Entwicklung kunftiger digitaler Verarbeitungssysteme. Fur Leistungsbewertungen dieser Systeme sind daher zunehmend flexible Analysemethoden und -Werkzeuge erforderlich, die moglichst systemunabhangig fur Leistungsbewertungen des Entwurfs, der Auswahl und von Modifikationen verwendbar sein sollten.

The Performance of Small Cache Memories in Minicomputer Systems with Several Processors

The evolution of the last years has shown that with improving technological possibilities, concepts of cache and hierarchical memories must be taken into consideration also for minicomputer systems [2, 8, 10]. As well-known, the system performance critically depends on placing the addressed data at the time of memory access in the cache, as the memory level being next to the processor. Therefore, size and structure of the cache, as well as the organizing strategies are of decisive importance [2, 3, 4, 7, 8].