A Comparison of Parallel Programming Models of Network Processors

Carsten Albrecht; Rainer Hagenau; Erik Maehle; Andreas Döring; Andreas Herkersdorf

A Comparison of Parallel Programming Models of Network Processors

Carsten Albrecht, Rainer Hagenau, Erik Maehle, Andreas Döring, Andreas Herkersdorf

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

5 Scopus citations

Abstract

Today’s network processor utilize parallel processing in order to cope with the traffic growth and wire-speed of current and future network technologies. In this paper, we study two important parallel programming models for network processors: run to completion and pipelining. In particular, the packet flow of a standard network application, IPv4 Forwarding, through two examined network processors, IBM PowerNP NP4GS3 and Intel IXP1200, is reviewed and characterized in respect to their programming models. Based on a benchmark for PC-cluster SANs, their application throughput and latency for Gigabit Ethernet is investigated and compared to a commercial, ASIC-based switch. It is shown that in this scenario network processors can compete with hard-wired solutions.

Original language	English
Title of host publication	ARCS 2004 - Organic and Pervasive Computing, Workshops Proceedings
Editors	Uwe Brinkschulte, Jurgen Becker, Dietmar Fey, Karl-Erwin Grosspietsch, Christian Hochberger, Erik Maehle, Thomas A. Runkler
Publisher	Gesellschaft fur Informatik (GI)
Pages	390-399
Number of pages	10
ISBN (Electronic)	3885793709
State	Published - 2004
Externally published	Yes
Event	17th Organic and Pervasive Computing Workshops, ARCS 2004 - Augsburg, Germany Duration: 26 Mar 2004 → …

Publication series

Name	Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI)
Volume	P-41
ISSN (Print)	1617-5468

Conference

Conference	17th Organic and Pervasive Computing Workshops, ARCS 2004
Country/Territory	Germany
City	Augsburg
Period	26/03/04 → …

Cite this

Albrecht, C., Hagenau, R., Maehle, E., Döring, A., & Herkersdorf, A. (2004). A Comparison of Parallel Programming Models of Network Processors. In U. Brinkschulte, J. Becker, D. Fey, K.-E. Grosspietsch, C. Hochberger, E. Maehle, & T. A. Runkler (Eds.), ARCS 2004 - Organic and Pervasive Computing, Workshops Proceedings (pp. 390-399). (Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI); Vol. P-41). Gesellschaft fur Informatik (GI).

Albrecht, Carsten ; Hagenau, Rainer ; Maehle, Erik et al. / A Comparison of Parallel Programming Models of Network Processors. ARCS 2004 - Organic and Pervasive Computing, Workshops Proceedings. editor / Uwe Brinkschulte ; Jurgen Becker ; Dietmar Fey ; Karl-Erwin Grosspietsch ; Christian Hochberger ; Erik Maehle ; Thomas A. Runkler. Gesellschaft fur Informatik (GI), 2004. pp. 390-399 (Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI)).

@inproceedings{7612a09724594e06a962efb5a1f305ac,

title = "A Comparison of Parallel Programming Models of Network Processors",

abstract = "Today{\textquoteright}s network processor utilize parallel processing in order to cope with the traffic growth and wire-speed of current and future network technologies. In this paper, we study two important parallel programming models for network processors: run to completion and pipelining. In particular, the packet flow of a standard network application, IPv4 Forwarding, through two examined network processors, IBM PowerNP NP4GS3 and Intel IXP1200, is reviewed and characterized in respect to their programming models. Based on a benchmark for PC-cluster SANs, their application throughput and latency for Gigabit Ethernet is investigated and compared to a commercial, ASIC-based switch. It is shown that in this scenario network processors can compete with hard-wired solutions.",

author = "Carsten Albrecht and Rainer Hagenau and Erik Maehle and Andreas D{\"o}ring and Andreas Herkersdorf",

year = "2004",

language = "English",

series = "Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI)",

publisher = "Gesellschaft fur Informatik (GI)",

pages = "390--399",

editor = "Uwe Brinkschulte and Jurgen Becker and Dietmar Fey and Karl-Erwin Grosspietsch and Christian Hochberger and Erik Maehle and Runkler, {Thomas A.}",

booktitle = "ARCS 2004 - Organic and Pervasive Computing, Workshops Proceedings",

}

Albrecht, C, Hagenau, R, Maehle, E, Döring, A & Herkersdorf, A 2004, A Comparison of Parallel Programming Models of Network Processors. in U Brinkschulte, J Becker, D Fey, K-E Grosspietsch, C Hochberger, E Maehle & TA Runkler (eds), ARCS 2004 - Organic and Pervasive Computing, Workshops Proceedings. Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI), vol. P-41, Gesellschaft fur Informatik (GI), pp. 390-399, 17th Organic and Pervasive Computing Workshops, ARCS 2004, Augsburg, Germany, 26/03/04.

A Comparison of Parallel Programming Models of Network Processors. / Albrecht, Carsten; Hagenau, Rainer; Maehle, Erik et al.
ARCS 2004 - Organic and Pervasive Computing, Workshops Proceedings. ed. / Uwe Brinkschulte; Jurgen Becker; Dietmar Fey; Karl-Erwin Grosspietsch; Christian Hochberger; Erik Maehle; Thomas A. Runkler. Gesellschaft fur Informatik (GI), 2004. p. 390-399 (Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI); Vol. P-41).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - A Comparison of Parallel Programming Models of Network Processors

AU - Albrecht, Carsten

AU - Hagenau, Rainer

AU - Maehle, Erik

AU - Döring, Andreas

AU - Herkersdorf, Andreas

PY - 2004

Y1 - 2004

N2 - Today’s network processor utilize parallel processing in order to cope with the traffic growth and wire-speed of current and future network technologies. In this paper, we study two important parallel programming models for network processors: run to completion and pipelining. In particular, the packet flow of a standard network application, IPv4 Forwarding, through two examined network processors, IBM PowerNP NP4GS3 and Intel IXP1200, is reviewed and characterized in respect to their programming models. Based on a benchmark for PC-cluster SANs, their application throughput and latency for Gigabit Ethernet is investigated and compared to a commercial, ASIC-based switch. It is shown that in this scenario network processors can compete with hard-wired solutions.

AB - Today’s network processor utilize parallel processing in order to cope with the traffic growth and wire-speed of current and future network technologies. In this paper, we study two important parallel programming models for network processors: run to completion and pipelining. In particular, the packet flow of a standard network application, IPv4 Forwarding, through two examined network processors, IBM PowerNP NP4GS3 and Intel IXP1200, is reviewed and characterized in respect to their programming models. Based on a benchmark for PC-cluster SANs, their application throughput and latency for Gigabit Ethernet is investigated and compared to a commercial, ASIC-based switch. It is shown that in this scenario network processors can compete with hard-wired solutions.

UR - http://www.scopus.com/inward/record.url?scp=33750210057&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:33750210057

T3 - Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI)

SP - 390

EP - 399

BT - ARCS 2004 - Organic and Pervasive Computing, Workshops Proceedings

A2 - Brinkschulte, Uwe

A2 - Becker, Jurgen

A2 - Fey, Dietmar

A2 - Grosspietsch, Karl-Erwin

A2 - Hochberger, Christian

A2 - Maehle, Erik

A2 - Runkler, Thomas A.

PB - Gesellschaft fur Informatik (GI)

T2 - 17th Organic and Pervasive Computing Workshops, ARCS 2004

Y2 - 26 March 2004

ER -

Albrecht C, Hagenau R, Maehle E, Döring A, Herkersdorf A. A Comparison of Parallel Programming Models of Network Processors. In Brinkschulte U, Becker J, Fey D, Grosspietsch KE, Hochberger C, Maehle E, Runkler TA, editors, ARCS 2004 - Organic and Pervasive Computing, Workshops Proceedings. Gesellschaft fur Informatik (GI). 2004. p. 390-399. (Lecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI)).

A Comparison of Parallel Programming Models of Network Processors

Abstract

Publication series

Conference

Other files and links

Fingerprint

Cite this