Abstract
The paper presents a new method for efficient fine grain computations on distributed memory computers. RDMA (Remote Direct Memory Access) communication is applied which assures direct access to memories of remote processing nodes. To obtain high RDMA efficiency for fine-grain computations with very frequent transmissions of small size messages, a specially designed structure of RDMA rotating buffers (RB) is introduced. It allows to fully exploit available communication bandwidth by provision of a special communication control infrastructure prepared and activated in a program before effective computations start. As an example of a fine-grain problem implemented with the RDMA rotating buffers, the discrete Fast Fourier Transform (FFT) execution is presented. “The binary-exchange algorithm” of FFT is examined showing efficiency of the RB method in comparison to standard MPI communication.
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
Tudruj, M., Masko, L.: Communication on the Fly and Program Execution Control in a System of Dynamically Configurable SMP Clusters. In: 11-th Euromicro Conference on Parallel Distributed and Network based Processing, Genoa - Italy, February 2003, pp. 67–74. IEEE Computer Society Press, Los Alamitos (2003)
Smyk, A., Tudruj, M.: OpenMP/MPI programming in a multi-cluster system based on shared memory/message passing communication. In: Grigoras, D., Nicolau, A., Toursel, B., Folliot, B. (eds.) IWCC 2001. LNCS, vol. 2326, pp. 241–248. Springer, Heidelberg (2001)
Tanabe, N., Hamada, Y., Nakajo, H., Imashiro, H., Yamamoto, J., Kudoh, T., Amano, H.: Low Latency Communication on DIMMnet-1 Network Interface Plugged into a DIMM Slot. In: International Conference in Parallel Computing in Electrical Engineering PARELEC 2002, Warsaw, September 22-25, pp. 9–14 (2002)
Hitachi Ltd: HI-UX/MPP - Remote DMA -C- User’s Guide Manual Number: 6A20- 3-021-10(E), Second Edition (January 1997)
Grama, A., Gupta, A., Karypis, G., Kumar, V.: Introduction to Parallel Computing, 2nd edn. Addison Vesley, Reading (2003)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2004 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Smyk, A., Tudruj, M. (2004). RDMA Communication Based on Rotating Buffers for Efficient Parallel Fine-Grain Computations. In: Wyrzykowski, R., Dongarra, J., Paprzycki, M., Waśniewski, J. (eds) Parallel Processing and Applied Mathematics. PPAM 2003. Lecture Notes in Computer Science, vol 3019. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-24669-5_7
Download citation
DOI: https://doi.org/10.1007/978-3-540-24669-5_7
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-21946-0
Online ISBN: 978-3-540-24669-5
eBook Packages: Springer Book Archive