Les reseaux de station de travail

Derniere mise a jour : 29/10/1998

English version of this page.

Introduction aux réseaux de station de travail

Copie des transparents présentés lors du tutoriel "Réseaux de stations de travail : alternative, avenir ou simple état transitoire des architectures parallèles extensibles" de la conférence RENPAR'8 (fichier postscript 2.6 Mo).

Les réseaux de station : les principaux pointeurs

Les réseaux de station "commerciaux"

L'IBM SPx

GIGAswitch et Memory Channel de DEC

Le Fujitsu AP3000

Des serveurs a base d'elements de grande diffusion

NCR WorldMark (SHVS : Standard High Volume Server d'Intel)

SEQUENT NUMA-Q (SHVS + SCI)

DATA GENERAL NUMA (SHVS + SCI)

Les principaux sites d'étude des réseaux de station de travail (classement alphabetique)

ATOMIC-2 (A l'origine de Myrinet, le pas suivant...)

AVALANCHE (HP + myrinet + unite de controle de cache et de communication)

GAMMA (12 Pentium sous linux + les messages actifs + Fast Ethernet)

HPVM (Clusters de Pentium + Myrinet, Servernet, etc.)

LANL Workstation Clusters (une etrangete : 8 RS/6000 + 1 SP2 (2noeuds) connectes par un Gigaswitch (dec))

MPC (reseau de PC)

NAS Davinci 8 SGI Power Challenge multiprocesseurs (quelques une monoprocesseur) connectees par Ethernet, HIPPI, FDDI et ATM (ouf!)

NOW (le projet federateur)

PAPERS (Reseau de stations avec operations globales n*n rapides)

PARASTATION (reseau specifique de stations et/ou PC)

RWCP PC Cluster (reseau Myrinet de 36 PC Pentium 166)

RWCP WCS (reseau Myrinet de 36 SuperSparc 20)

SHRIMP (reutilisation d'un reseau de Paragon)

StarT-Voyager (clusters de PPC 604, NES : Network Endpoint Subsystem et Arctic router)

DSM Projects (Memoire virtuelle partagee)

Wisconsin Wind Tunnel (COW, Vortex, Typhoon, Tampest, etc.)

DMRC (Distributed Multimedia Research Center)

Les principaux réseaux rapides

MYRINET

ATM (FORE)

SCI (Vitesse, Dolphin)

SERVERNET (un reseau de TANDEM destine a la grande diffusion, anciennement TNET)

Les principales API rapides

VIA: Virtual Interface Architecture

MESSAGE ACTIF

FAST MESSAGE

U-Net

BIP(Messagerie tres rapide pour Myrinet et x86)

Des modèles de programmation pour réseau de multiprocesseurs

SIMPLE (Primitives de communication collective)

Des modèles de programmation pour ressources à disponibilité variable

MPVM (passage de messages et migration de processus)

NEXUS (multithreading)

PM2 (multithreading)

Cilk (multithreading)

Programmation objet

CORBA

JavaRMI

Des systèmes d'exploitation pour réseaux de station de travail

AMOEBA (système micro-noyau distribué)

Chorus (système micro-noyau)

GLUNIX (surcouche d'UNIX)

MOSIX (surcouche d'UNIX)

Des gestionnaires de ressources pour réseaux de station de travail

Cluster Management Software (tour d'horizon sur les principaux gestionnaires)

Condor (un classique du genre)

LoadLeveler (solution commercial d'IBM)

Des systémes de fichiers

xFS (systéme de fichiers du projet NOW)

Des reseaux de station de travail utilises en production

RTCPP (Regional Training Center for Parallel Processing, North Carolina State University)

NAS Report (McDonnell Douglas, Pratt & Whitney, Boeing)

PEIRS (Parallel Integrated Environment for Reliability of Structures of existing industrial plants)

Conferences/Workshops

Second NOW/Cluster Workshop

Autres listes de pointeurs sur les reseaux de station

ASPLOS Workshop NOW/COW

AmesLab workshop

Quelques publications sur le sujet :

[AMZ96] C. Amza et. al., "TreadMarks: Shared Memory Computing on Networks of Workstations", IEEE Computer, Feb 1996.

[AND95] T. E. Anderson et. al., "A case for NOW (Networks of Workstations)", IEEE Micro, Feb 1995.

[AND96] T. E. Anderson et. al., "Serveless Network File Systems", ACM Tran. Computer Systems, Vol 14, No. 1, February 1996.

[ARP 94] R. H. Arpaci et. al., "The Interaction of Parallel and Sequential Workloads on a Network of Workstations", UC Berkeley Technical Report CS-94-838, 1994, http://now.cs.berkeley.edu

[ASA95] S. Asami et. al., "The Design of Large-Scale, Do-It-Yourself RAIDs", submited but not yet published, draft available at http://now.cs.berkeley.edu, Nov 1995.

[BER93] B. Bergten, M, Couprie and P. Valduriez, "Overview of Parallel Architectures for Databases", The computer journal, Vol. 36, No. 8, 1993.

[BLU95] M. A. Blumrich et. al., "Virtual-Memory-Mapped Network Interfaces", IEEE Micro, Feb 1995.

[BOD95] N. J. Boden et. al., "Myrinet: A Gigabit-per-Second Local Area Network", IEEE Micro, Feb 1995.

[CAS95] J. Casas et. al., "MPVM: A Migration Transparent Version of PVM", Computing Systems, Vol. 8, No. 2, Spr 1995.

[CHA96] C.-C. Chang et. al., "Design and Performance of Active Messages on the SP-2", Cornell CS Tech. Report 96-1572, Feb 1996.

[COX94] A. L. Cox et. al., "Software Versus Hardware Shared-Memory Implementation: A case Study", Proceedings of International Symposium on Computer Architecture, 1994.

[CVE96] Z. Cvetanovic and D. Bhandarkar, "Performance characterization of the Alpha 21164 Microprocessor using TP and SPEC workloads", Proceedings of second HPCA,1996.

[DAH94] M. D. Dahlin et. al., "Cooperative Caching: Using Remote Client Memory to Improve File System Performance", Proceedings of OSDIÕ94, 1994.

[EDJ94] G. Edjlali et. al., "DataParallel Programming in An Adaptative Environment", Tech Report CS-TR-3350 and UMIACS-TR-94-109, University of Maryland, DCS and UMIACS, Sept 1994.

[EIK92] T von Eicken et. al., "Active Messages: A Mechanism for Integrated Communication and Computation", Proceedings of 19th ISCA, May 1992.

[EIK95] T. von Eicken et. al., "Low-Latency Communication Over ATM Networks Using Active Messages", IEEE Micro, Feb 1995.

[GIL96] R. B. Gillett, "Memory Channel Network for PCI", IEEE Micro, Feb 1996.

[GUS92] D. B. Gustavson, "The Scalable Coherent Interface and the Related Standards Projects", IEEE Micro 12, Feb 1992.

[GUS94] D. Gustavson and A. Li, "Local Area MultiProcessor: The Scalable Coherent Interface", in Defining the Global Information Infrastructure: Infrastructure, Systems, and Services, Stephen F. Lundstrom, Editor, SPIE Press, Vol 56, 1994.

[GUS95] D. Gustavson and Q. Li, "SCI-based Local Area MultiProcessor (LAMP)", Tutorial Notes of HPCA 1, Jan 1995.

[HAL96] T. R. Halfhill, "Inside the Web PC", Byte, Mar 1996.

[HOR95] R. W. Horst, "TNet: A Reliable System Area Network", IEEE Micro, Feb 1995.

[KAR96] M. Karlsson and P. Stenstršm, "Performance of a Cluster-Based Multiprocessor Built from ATM Switches and Bus-Based Multiprocessor Servers", Prpceedings of HPCA II, Feb 1996.

[KEL94] P. Keleher et. al., "TreadMarks: Distributed Shared Memory on Standard Workstations and Operating Systems", Proceedings of Usenix Conference, Winter 94, 1994.

[MIN95] R. Minnich, D. Burns, F. Hady, "The Memory-Integrated Network Interface", IEEE Micro, Feb 1995.

[MUT91] M. W. Mutka, M. Livny, "The available capacity of a privately owned workstation environment", Performance Evaluation 12, Elsevier, North-Holland, 1991

[MUT92] M. W. Mutka, "Estimating Capacity For Sharing in a Privately Owned Workstation Environment", IEEE Trans. Software Engineering, Vol 18, No. 4, April 1992.

[PAK95] Scott Pakin et. al., "High Performance Messaging on Workstations: Illinois Fast Messages (FM) for Myrinet, Proceedings of SupercomputingÕ 95, 1995.

[REN89] R. Van Renesse, H. Van Staveren and A. S. Tanenbaum, "The Performance Of The Amoeba Distributed Operating System", Software-Practice and Experience, Vol 19, March 1989.

[SCO96] S. Scott, "The Gigaring Channel", IEEE Micro, Vol 16, No. 1, Feb 1996.

[TAN91] A. S. Tanenbaum, M. F. Kaashoek, R. van Renesse, H. Bal, "The Amoeba Distributed Operating System - A Status Report", Computer communications, Vol. 14 , pp. 324-335, Jul 1991.

[WAL94] D. W. Walker, "The design of a standard message passing interface for distributed memory concurrent computers", Parallel Computing, Vol 20, 1994.

[WOO95] Alan Wood, "Predicting Client/Serveur Availability", IEEE Computer, Apr 1995.

Contact

Pour toutes informations complémentaires, remarques, discussions, proposition de sites à ajouter dans la liste, contacter : Franck Cappello