A Distributed Implementation of Flat Concurrent Prolog on by Uwe Glässer PDF

By Uwe Glässer

ISBN-10: 3322976114

ISBN-13: 9783322976116

ISBN-10: 3322976122

ISBN-13: 9783322976123

Show description

Read or Download A Distributed Implementation of Flat Concurrent Prolog on Message-Passing Multiprocessor Systems PDF

Similar german_5 books

Parallele adaptive Mehrgitterverfahren by Peter Bastian PDF

Aus dem Inhalt Komponenten adaptiver Verfahren - Parallelisierung des Mehrgitterverfahrens - Parallele Gitterverfeinerung - Der Programmbaukasten UG - Dynamische Lastverteilung für adaptive Mehrgitterverfahren - Praktische Ergebnisse und Ausblick - Anhang: Übersicht über Lastverteilungsverfahren

Download e-book for iPad: Höhere Mathematik für Ingenieure: Band III: Gewöhnliche by Klemens Burg, Herbert Haf, Friedrich Wille, Andreas Meister

Das Buch ist Teil einer Vorlesungsreihe, die sich über die ersten vier bis fünf Semester erstreckt. Es wendet sich in erster Linie an Studierende der Ingenieurwissenschaften, darüber hinaus aber allgemein an Studierende aller technischer und physikalischer Fachrichtungen sowie an Studierende der Angewandten Mathematik.

Get Wie Führe Ich Richtig?: Personalführung mit Autorität PDF

Es wird die Notwendigkeit der Autorit? t in der Personalf? hrung beschrieben und wie diese Autorit? t im positiven Sinn zum Vorteil von Unternehmen, Vorgesetzten und Mitarbeitern in humaner Weise ausge? bt werden kann. Viele Checklisten f? hren den Vorgesetzten zu richtigem Verhalten.

Extra resources for A Distributed Implementation of Flat Concurrent Prolog on Message-Passing Multiprocessor Systems

Sample text

Each of these reduction attempts may require to compute a number of resulting variable bindings. Whenever an attempt to reduce a process by means of a selected dause fails, the obtained variable bindings have to be undone as such an attempt must not leave any trace. cHT"uckiug. When a process reduction is delayed depending on the instantiation of one or more read-only variables, dause selection may become even more costly. A process which is scheduled several times, but cannot be reduced because of uninstantiated read-only variables, requires to evaluate the whole program procedure more than once.

In fact, this means that the programmer need not to know how many processors are involved in the execution of a program and how the global process network has to be mapped onto these processors. The same application program may run on a uniprocessor configuration as weil as on arbitrary large networks with any nummer of processors. Consequently, the compiler needs also not to know whether the code it produces is to be executed on a sequential or parallel machine. On the other hand, it still remains possible to extract helpful information from a program at compile-time in order to support or improve parallelization whenever the program is executed by more than one processor.

In the design and implementation of our machine we use a different. instruction set as weIl as different representations for FCP data structures. Neverthdess, there are certain similarities with respect to basic design concepts. ailed descriptions including the applied instruction set for the original and an improved version of this machine are presented in [Lehrenfeld90] and [Hannesen91], rcspcctivcly. ion, a further improved machine architecture was investigated in [GruIlzig92]. 2 CHAPTER :l.

Download PDF sample

A Distributed Implementation of Flat Concurrent Prolog on Message-Passing Multiprocessor Systems by Uwe Glässer


by Charles
4.3

Rated 4.97 of 5 – based on 25 votes