[ News | IPCA | Mirrors | Add | Search | Mail | Help | WoTUG ]
Abstract: This paper presents an overview of PUMA, Performance-oriented, User-managed Messaging Architecture, a message passing kernel. Message passing in PUMA is based on portals - an opening in the address space of an application process. Once an application process has established a portal, other processes can write values into the portal using a simple send operation. Because messages are written directly into the address space of the receiving process, there is no need to buffer messages in the PUMA kernel and later copy them into the applications address space. PUMA consists of two components: the quintessential kernel Q-Kernel and the process control thread PCT. While the PCT provides management decisions, the Q-Kernel controls access and implements the policies specified by the PCT.
Authors: Stephen R. Wheat; Arthur B. Maccabe; Rolf Riesen; David W. van Dresser and T. Mack Stallcup.
Abstract: An implementation of the LINPACK benchmark is described which achieves 72.9 Gflop/sec on 1872 nodes of an Intel Paragon. Implications concerning the architecture of the Paragon and the necessity of a high performance operating system like SUNMOS.
Authors: David Womble; David Greenberg; Stephen Wheat and Rolf Riesen.
Abstract: This booklet is written in the hope that we can accelerate the learning process of readers using an Intel Paragon computer for the first time. It is not intended as a replacement for the manufacturers programming and reference manuals, but rather to complement them and to guide the new user through the jungle of terminology and programming techniques. We hope that we have written this guide such that the reader can complete this tutorial without requiring such manuals. Our long experience with advanced computer systems has shown us that these users want to become familiar with a new system as quickly as possible, without having to wade through, say, a metre shelf of manuals to be able to write a simple program. Rather than write yet another tutorial, we have tried to give the user guidance in using the Paragon, with concise information, with the hope that they can get up to speed as quickly as possible. Very little instruction on parallel algorithms is given in this guide, we have concentrated instead on use of the system. The text and examples in this guide were developed on the 100 node system at parallab. Almost everything in this guide should be applicable to all Paragon system configurations.
COPYRIGHT NOTICE: This document may be distributed freely for educational purposes provided it is distributed in its entirety, including this copyright notice. This booklet may be used as course material or as on-line documentation provided acknowledgement is given to parallab . If copies of this document are distributed then parallab must be informed at the address below: parallab secretary, parallab, Dept. of Informatics, University of Bergen, N-5020 Bergen, Norway email: adm@parallab.uib.no.
Author: Jeremy Cook, Parallab, Dept. of Informatics, University of Bergen, N-5020 Bergen, Norway