Abstract
In the state-of-the-art production quality MPI (Message Passing Interface) libraries, communication progress is either performed by the main thread or a separate communication progress thread. Taking advantage of separate communication threads can lead to a higher overlap of communication and computation as well as reduced total application execution time. However, such an approach can also lead to contention for CPU resources leading to sub-par application performance as the application itself has less number of available cores for computation. Recently, Mellanox has introduced the BlueField series of adapters which combine the advanced capabilities of traditional ASIC based network adapters with an array of ARM processors. In this paper, we propose BluesMPI, a high performance MPI non-blocking Alltoall design that can be used to offload MPI_Ialltoall collective operations from the host CPU to the Smart NIC. BluesMPI guarantees the full overlap of communication and computation for Alltoall collective operations while providing on-par pure communication latency to CPU based on-loading designs. We explore several designs to achieve the best pure communication latency for MPI_Ialltoall. Our experiments show that BluesMPI can improve the total execution time of the OSU Micro Benchmark for MPI_Ialltoall and P3DFFT application up to 44% and 30%, respectively. To the best of our knowledge, this is the first design that efficiently takes advantage of modern BlueField Smart NICs in deriving the MPI Alltoall collective operation to get peak overlap of communication and computation.
This research is supported in part by National Science Foundation grants #1818253, #1854828, #1931537, #2007991, #2018627, and XRAC grant #NCR-130002.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
High-Performance Center Overview. https://www.hpcadvisorycouncil.com/cluster_center.php
Mellanox BlueField. https://docs.mellanox.com/x/iQO3
Panda, D.K., Subramoni, H., Chu, C.H., Bayatpour, M.: The MVAPICH project: Transforming research into high-performance MPI library for HPC community. J. Comput. Sci. 101208 (2020)
Bayatpour, M., et al.: Communication-aware hardware-assisted MPI overlap engine. In: Sadayappan, P., Chamberlain, B.L., Juckeland, G., Ltaief, H. (eds.) ISC High Performance 2020. LNCS, vol. 12151, pp. 517–535. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-50743-5_26
Bayatpour, M., Ghazimirsaeed, S.M., Xu, S., Subramoni, H., Panda, D.K.: Design and characterization of infiniband hardware tag matching in MPI. In: 20th Annual IEEE/ACM CCGRID (2020)
Gropp, W., Lusk, E., Doss, N., Skjellum, A.: A high-performance, portable implementation of the MPI, message passing interface standard. Technical report, Argonne National Laboratory and Mississippi State University
InfiniBand Trade Association (2017). http://www.infinibandta.com
Kandalla, K., Subramoni, H., Tomko, K., Pekurovsky, D., Sur, S., Panda, D.K.: High-performance and scalable non-blocking all-to-all with collective offload on infiniband clusters: a study with parallel 3D FFT. Comput. Sci. 26, 237–246 (2011)
Liu, M., Cui, T., Schuh, H., Krishnamurthy, A., Peter, S., Gupta, K.: iPipe: a framework for building distributed applications on SmartNICs. In: SIGCOMM 2019: Proceedings of the ACM Special Interest Group on Data Communication, pp. 318–333 (2019). https://doi.org/10.1145/3341302.3342079
Network-Based Computing Laboratory: MVAPICH2-X (Unified MPI+PGAS Communication Runtime over OpenFabrics/Gen2 for Exascale Systems). http://mvapich.cse.ohio-state.edu/overview/mvapich2x/
Pekurovsky, D.: P3DFFT library (2006–2009). www.sdsc.edu/us/resources/p3dfft/
Phothilimthana, P.M., Liu, M., Kaufmann, A., Simon Peter, R.B., Anderson, T.: Floem: a programming system for NIC-accelerated network applications. In: Proceedings of the 2010 IEEE 26th Symposium on Mass Storage Systems and Technologies (MSST). 13th USENIX Symposium on Operating Systems Design and Implementation (2018)
Potluri, S., et al.: MVAPICH-PRISM: a proxy-based communication framework using InfiniBand and SCIF for Intel MIC clusters. In: Proceedings of SC 2013, SC 2013, pp. 54:1–54:11 (2013)
Scalable hierarchical aggregation protocol: scalable hierarchical aggregation protocol. https://www.mellanox.com/products/sharp
Subramoni, H., Kandalla, K., Sur, S., Panda, D.K.: Design and evaluation of generalized collective communication primitives with overlap using ConnectX-2 offload engine. In: Internationall Symposium on Hot Interconnects (HotI), August 2010 (2010)
Subramoni, H., et al.: Designing non-blocking personalized collectives with near perfect overlap for RDMA-enabled clusters. In: Kunkel, J.M., Ludwig, T. (eds.) ISC High Performance 2015. LNCS, vol. 9137, pp. 434–453. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-20119-1_31
Sur, S., Jin, H.W., Chai, L., Panda, D.K.: RDMA read based rendezvous protocol for MPI over infiniband: design alternatives and benefits. In: Proceedings of the Eleventh ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPoPP 2006 (2006)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Bayatpour, M., Sarkauskas, N., Subramoni, H., Maqbool Hashmi, J., Panda, D.K. (2021). BluesMPI: Efficient MPI Non-blocking Alltoall Offloading Designs on Modern BlueField Smart NICs. In: Chamberlain, B.L., Varbanescu, AL., Ltaief, H., Luszczek, P. (eds) High Performance Computing. ISC High Performance 2021. Lecture Notes in Computer Science(), vol 12728. Springer, Cham. https://doi.org/10.1007/978-3-030-78713-4_2
Download citation
DOI: https://doi.org/10.1007/978-3-030-78713-4_2
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-78712-7
Online ISBN: 978-3-030-78713-4
eBook Packages: Computer ScienceComputer Science (R0)