Abstract
The overall performance of networking computation is determined by the grid node computer and the network infrastructure under it. But the grid node computer currently in use is not designed for grid applications. At same time, router capacity is getting more and more powerful. To Combine high performance router with traditional supercomputer will be a feasible way for both supercomputing and high performance QoS routing. At one side, it can release the node computer from the networking overhead and focus on computing and data processing, At the other side, it bring more powerful computing power to router and enable it to meet stream-based QoS and security processing requirement. This paper proposed an innovative integrated architecture AQRGS(Architecture for QoS-enable Router and Grid–oriented Supercomputer), which makes the CPUs focus on their high performance computing and data processing and leave the network communication jobs to the more specialized network processors in the high performance router. Thus the computing and communication capability of the grid node computers will be both highly improved and the grid application will run much faster. This paper discussed the related issues of building a grid oriented parallel computer integrating the network processing, which can also be benefit for QoS processing when used as a high performance router.
This work is supported by National Natural Science Foundation of China (NSFC), under agreement no 90104001, National Basic Research Priorities Program of China, under agreement no.2003CB314802.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Ali, M., Chiruvolu, G., Ge, A.: Traffic Engineering in Metro Ethernet. IEEE Network Magazine, Alcatel (March/April 2005)
Knutsson, B., Lu, H., Xu, W., Hopkins, B.: Peer-to-Peer Support for Massively Multiplayer Games. In: Infocom 2004 (2004)
Ulme, C., Wood, C., Yalamanchili, S.: Active SANs: Hardware Support for Integrating Computation and Communication. In: Proceedings of the Workshop on Novel Uses of System Area Networks at HPCA, SAN 2002 (2002)
Boden, N., Cohen, D., Felderman, R., Kulawik, A., Seitz, C., Seizovic, J., Su, W.: Myrinet: A Gigabit-per-second Local Area Network. IEEE Micro. 15(1) (1995)
Eichen, T., Culler, D., Goldstein, S., Schauser, K.: Active Messages: A Mechanism for Integrated Communication and Computation. In: Proceedings of The 19th Annual International Symposium on Computer Architecture, pp. 256–266 (1992)
Tennenhouse, D.L., Smith, J.M., Sincoskie, W.D., et al.: A survey of active network research. IEEE Communication Magazine 35(1), 80–86 (1997)
Buntinas, D., Duato, J., Sadayappan, P., Panda, D.K.: NIC-Assisted Broadcast/Multicast on Myrinet. In: Workshop on Communication and Architectural Support for Network-Based Parallel Computing (CANPC) (January 2000)
Panda, D.K.: Active Network Interface: Opportunities and Challenges. In: Proceedings of 27th Annual IEEE Conference on Local Computer Networks (LCN 2002), Tampa, FL, USA, November 6-8, pp. 605–608. IEEE Computer Society, Los Alamitos (2002) ISBN 0-7695-1591-6
Bosilca, G., Bouteiller, A., Cappello, F., et al.: MPICH-V: Toward a Scalable Fault Tolerant MPI for Volatile Nodes. In: SC 2002 (2002)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2005 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Wu, C., Yang, X. (2005). An Integrated Architecture for QoS-Enable Router and Grid-Oriented Supercomputer. In: Lu, X., Zhao, W. (eds) Networking and Mobile Computing. ICCNMC 2005. Lecture Notes in Computer Science, vol 3619. Springer, Berlin, Heidelberg. https://doi.org/10.1007/11534310_127
Download citation
DOI: https://doi.org/10.1007/11534310_127
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-28102-3
Online ISBN: 978-3-540-31868-2
eBook Packages: Computer ScienceComputer Science (R0)