Abstract
Existing multithreaded network processors architecture with multiple processing engines (PEs), aims at taking advantage of blocked multithreading technique which executes instructions of different user-defined threads in the same PE pipeline, in explicit and interleave way. Multiple PEs, each of which is a multithreaded processor core, process several packets in parallel to hide long memory access latency. Most of them are optimized for throughputs mostly in data-plane. In future network workloads, the boundaries between data-plane and control-plane become blurred, so that PEs are demanded not only wire speed packet forwarding on data-plane, but also highly intelligent and increased complex packet processing function on control-plane. In this paper, we analyze SMT’s short latency tolerance potential when used in out-of-order and dynamic scheduling PE cores. We show in this paper that 2~4 issue SMT provides an excellent short memory and branch latency tolerance, which gain higher instructions throughout as well as much simpler structures.
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
Intel Corporation: Intel IXP2400 Network Processor Family Hardware Reference Manual (June 2001)
IBM Corporation: The Network Processor: Enabling Technology for High-Performance Networking. IBM Microelectronics (1999)
C-Port Corporation: C-5 Digital Communications Processor (1999), http://www.cportcorp.com/solutions/docs/c5brief.pdf
Wolf, T., Franklin, M.: CommBench - A Telecommunications Benchmark for Network Processors. In: International Symposium on Performance Analysis of Systems and Software (April 2000)
Memik, G., Mangione-smith, W., Hu, W.: NetBench: A Benchmarking Suite for Network Processors. In: 2001 IEEE/ACM International Conference on Computer-Aided Design (2001)
Lee, B.K.: NpBench: A Benchmark Suite for Control plane and Data plane Applications for Network Processors. In: IEEE International Conference on Computer Design (October 2003)
Gonçalves, R., Ayguadé, E., Valero, M., Navaux, P.: A Simulator for SMT Architectures: Evaluating Instruction Cache Topologies. In: SBAC-PAD, Brazil, pp. 2169–2161 (2000)
Simplescalar Simulator, from, http://www.simplerscalar.com
Chiueh, T.-C., Pradhan, P.: Cache Memory Design for Network Processors. In: Proceeding of the 6th International Symposium. on High Performance Computer Architecture, Tolouse, France (January 2000)
Sherwood, T., Varghese, G., Calder, B.: A Pipelined Memory Architecture for High Throughput Network Processors. In: Proceedings of the 30th Annual, ISCA 2003 (2003)
Hasan, J., Chandra1, S., Vijaykumar, T.N.: Efficient Use of Memory Bandwidth to Improve Network Processor Throughput. In: Proceedings of the 30th Annual, ISCA 2003 (2003)
Parcerisa, J.-M., Gonzalez, A.: Improving Latency Tolerance of Multithreading through Decoupling. IEEE Transactions on Computers 50(10) (October 2001)
Hily, S., Seznec, A.: Branch Prediction and Simultaneous Multithreading. In: proceeding of International Conference on Parallel Architecture and Compilation Techniques (1996)
Ramsay, M., Feucht, C., Lipasti, M.H.: Exploring Efficient SMT Branch Predictor Design. In: Workshop on Complexity-Effective Design, in conjunction with ISCA (June 2003)
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
Liang, B., An, H., Lu, F., Guo, R. (2005). Improving Latency Tolerance of Network Processors Through Simultaneous Multithreading. In: Cao, J., Nejdl, W., Xu, M. (eds) Advanced Parallel Processing Technologies. APPT 2005. Lecture Notes in Computer Science, vol 3756. Springer, Berlin, Heidelberg. https://doi.org/10.1007/11573937_9
Download citation
DOI: https://doi.org/10.1007/11573937_9
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-29639-3
Online ISBN: 978-3-540-32107-1
eBook Packages: Computer ScienceComputer Science (R0)