{"title":"Advancing application process affinity experimentation: open MPI's LAMA-based affinity interface","authors":"Joshua Hursey, J. Squyres","doi":"10.1145/2488551.2488603","DOIUrl":"https://doi.org/10.1145/2488551.2488603","url":null,"abstract":"Application studies have shown that the tuning of Message Passing Interface (MPI) processes placement in a server's non-uniform memory access (NUMA) networking topology can have a dramatic impact on performance. The performance implications are magnified when running a parallel job across multiple server nodes, especially with large scale MPI applications. As processor and NUMA topologies continue to grow more complex to meet the demands of ever-increasing processor core counts, best practices regarding process placement also need to evolve.\u0000 This paper presents Open MPI's flexible interface for distributing the individual processes of a parallel job across processing resources in a High Performance Computing (HPC) system, paying particular attention to the internal server NUMA topologies. The interface is a realization of the Locality-Aware Mapping Algorithm (LAMA) [8], and provides both simple and complex mechanisms for specifying regular process-to-processor mappings and affinitization. Open MPI's LAMA implementation is intended as a tool for MPI users to experiment with different process placement strategies on both current and emerging HPC platforms.","PeriodicalId":243229,"journal":{"name":"European MPI Users Group Meeting","volume":"302 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115439942","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Gonzalo Martín, D. E. Singh, M. Marinescu, J. Carretero
{"title":"Parallel algorithm for simulating the spatial transmission of influenza in EpiGraph","authors":"Gonzalo Martín, D. E. Singh, M. Marinescu, J. Carretero","doi":"10.1145/2488551.2488585","DOIUrl":"https://doi.org/10.1145/2488551.2488585","url":null,"abstract":"This paper introduces an approach to modeling and simulating the propagation of flu-like infectious diseases over large, widely spread urban areas connected by transportation networks. We incorporate geographic location and a transportation model into our region-based, closed-world EpiGraph simulator to realistically model the movement of the virus between different geographic regions. The resulting simulator can assist in understanding how outbreaks propagate between far apart regions due to the movement of people outside their base location. This paper describes the MPI-based implementation of EpiGraph and its performance evaluation when simulating large-scale scenarios. We evaluate the simulator both on a distributed memory system and on a shared memory system.","PeriodicalId":243229,"journal":{"name":"European MPI Users Group Meeting","volume":"97 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125084398","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Amin Hassani, A. Skjellum, R. Brightwell, Brian W. Barrett
{"title":"Design, implementation, and performance evaluation of MPI 3.0 on portals 4.0","authors":"Amin Hassani, A. Skjellum, R. Brightwell, Brian W. Barrett","doi":"10.1145/2488551.2488563","DOIUrl":"https://doi.org/10.1145/2488551.2488563","url":null,"abstract":"The latest version of the Portals interconnect programming interface contains several improvements intended for MPI point-to-point and one-sided communication operations, including new functionality defined in MPI-3. This paper discusses the rationale for these improvements to Portals and describes how they can be used in an MPI implementation. We provide preliminary micro-benchmark performance results using a reference implementation of Portals over InfiniBand Verbs and the Open MPI implementation.","PeriodicalId":243229,"journal":{"name":"European MPI Users Group Meeting","volume":"64 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126769959","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Optimization of MPI persistent communication","authors":"M. Hatanaka, A. Hori, Y. Ishikawa","doi":"10.1145/2488551.2488566","DOIUrl":"https://doi.org/10.1145/2488551.2488566","url":null,"abstract":"This paper proposes a novel optimization technique for MPI persistent communication, that utilizes multiple RDMA engines to carry out low latency communication. Because the interconnects used in modern supercomputers have multiple RDMA engines, the multiple communication requests specified by a persistent communication invocation can be scheduled onto the RDMA engines in an optimal way and thus result in better communication performance. Such a scheduling algorithm is not only a packing problem, but also avoids interconnect resource contentions as much as possible. The scheduling algorithm proposed in this paper balances load of RDMA engines and mitigates network link contentions in case of neighbor communication patterns such like in stencil computation. The proposed scheduling algorithm is implemented in Open MPI of K computer using RDMA functions provided as Fujitsu MPI extensions. A typical 2D stencil computation of a climate simulation code is used as a benchmark program. The experimental result shows that a factor of two speedup of communication time is achieved.","PeriodicalId":243229,"journal":{"name":"European MPI Users Group Meeting","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115574874","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Parallelizing a multiobjective swarm intelligence approach to phylogenetics using hybrid MPI/OpenMP schemes","authors":"Sergio Santander-Jiménez, M. A. Vega-Rodríguez","doi":"10.1145/2488551.2488576","DOIUrl":"https://doi.org/10.1145/2488551.2488576","url":null,"abstract":"Phylogenetic inference is one of the most challenging problems in Computational Biology. As recent research lines aim to introduce multiobjective optimization techniques to resolve incongruences in Phylogenetics, parallel multiobjective metaheuristics can be useful to address the computational complexity required to perform phylogenetic analyses according to multiple criteria simultaneously. In this work, we propose several master-worker hybrid approaches based on MPI and OpenMP to parallelize a multiobjective algorithm inspired by the behaviour of fireflies for inferring phylogenies on multicore cluster architectures. Experiments on four real biological data sets suggest that this algorithm can achieve significant speedup and efficiency values by using a proper hybrid model designed to exploit parallelism at the inference and assessment levels.","PeriodicalId":243229,"journal":{"name":"European MPI Users Group Meeting","volume":"10 3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126684270","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Masamichi Takagi, Yuichi Nakamura, A. Hori, Balazs Gerofi, Y. Ishikawa
{"title":"Revisiting rendezvous protocols in the context of RDMA-capable host channel adapters and many-core processors","authors":"Masamichi Takagi, Yuichi Nakamura, A. Hori, Balazs Gerofi, Y. Ishikawa","doi":"10.1145/2488551.2488571","DOIUrl":"https://doi.org/10.1145/2488551.2488571","url":null,"abstract":"We revisit RDMA-based rendezvous protocols in MPI in the context of cluster computer with RDMA-capable HCA and many-core processors, and propose two improved protocols. The conventional sender-initiate rendezvous protocols cause costly processor-device communications via PCI bus on detecting completion of RDMA transfer. The conventional receiver-initiate rendezvous protocols need to send extra control messages when a value of the memory-slot to poll in the receive buffer has the same value as the send buffer. The first proposed protocol implements polling on a memory-slot in the receive buffer to eliminate the processor-device communications. The second proposed protocol randomizes the value of the memory-slot to poll to reduce extra control messages. We have evaluated the proposed protocols using micro-benchmarks and NAS Parallel Benchmarks. One of the proposed protocols has a benefit compared to the conventional protocols. And the second proposed protocol reduces the execution time by up to 11.14% compared to the first protocol.","PeriodicalId":243229,"journal":{"name":"European MPI Users Group Meeting","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133472107","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Christi Symeonidou, Polyvios Pratikakis, A. Bilas, Dimitrios S. Nikolopoulos
{"title":"DRASync: distributed region-based memory allocation and synchronization","authors":"Christi Symeonidou, Polyvios Pratikakis, A. Bilas, Dimitrios S. Nikolopoulos","doi":"10.1145/2488551.2488558","DOIUrl":"https://doi.org/10.1145/2488551.2488558","url":null,"abstract":"We present DRASync, a region-based allocator that implements a global address space abstraction for MPI programs with pointer-based data structures. The main features of DRASync are: (a) it amortizes communication among nodes to allow efficient parallel allocation in a global address space; (b) it takes advantage of bulk deallocation and good locality with pointer-based data structures. (c) it supports ownership semantics of regions by nodes akin to reader-writer locks, which makes for a high-level, intuitive synchronization tool in MPI programs, without sacrificing message-passing performance. We evaluate DRASync against a state-of-the-art distributed allocator and find that it produces comparable performance while offering a higher level abstraction to programmers.","PeriodicalId":243229,"journal":{"name":"European MPI Users Group Meeting","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114499620","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
S. Sehrish, S. Son, W. Liao, A. Choudhary, K. Schuchardt
{"title":"Improving collective I/O performance by pipelining request aggregation and file access","authors":"S. Sehrish, S. Son, W. Liao, A. Choudhary, K. Schuchardt","doi":"10.1145/2488551.2488559","DOIUrl":"https://doi.org/10.1145/2488551.2488559","url":null,"abstract":"In this paper, we propose a multi-buffer pipelining approach to improve collective I/O performance by overlapping the dominant request aggregation phases with the I/O phase in the two-phase I/O implementation. Our pipelining method first divides the collective buffer into a group of small size buffers for an individual collective I/O call and then pipelines the asynchronous communication to exchange the I/O requests with the I/O requests sent to the file system. Our performance evaluation of a representative I/O benchmark and a production application shows 20% improvement in the I/O time, given theoretical upper bound of 50% when both phases completely overlap.","PeriodicalId":243229,"journal":{"name":"European MPI Users Group Meeting","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123668512","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Marc-André Hermanns, Manfred Miklosch, David Böhme, F. Wolf
{"title":"Understanding the formation of wait states in applications with one-sided communication","authors":"Marc-André Hermanns, Manfred Miklosch, David Böhme, F. Wolf","doi":"10.1145/2488551.2488569","DOIUrl":"https://doi.org/10.1145/2488551.2488569","url":null,"abstract":"To better understand the formation of wait states in MPI programs and to support the user in finding optimization targets in the case of load imbalance, a major source of wait states, we added in our earlier work two new trace-analysis techniques to Scalasca, a performance analysis tool designed for large-scale applications. In this paper, we show how the two techniques, which were originally restricted to two-sided and collective MPI communication, are extended to cover also one-sided communication. We demonstrate our experiences with benchmark programs and a mini-application representing the core of the POP ocean model.","PeriodicalId":243229,"journal":{"name":"European MPI Users Group Meeting","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125899992","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Optimization of MPI_Allreduce on the blue Gene/Q supercomputer","authors":"Sameer Kumar, Daniel Faraj","doi":"10.1145/2488551.2488557","DOIUrl":"https://doi.org/10.1145/2488551.2488557","url":null,"abstract":"The IBM Blue Gene/Q supercomputer has a 5D torus network where each node is connected to ten bi-directional links. In this paper we present techniques to optimize the MPI_Allreduce collective operation by building ten different edge disjoint spanning trees on the ten torus links. We accelerate summing of network packets with local buffers by the use of Quad Processing SIMD unit in the BG/Q cores and executing the sums on multiple communication threads created by the PAMI libraries. The net gain we achieve is a peak throughput of 6.3 GB/sec for double precision floating point sum allreduce, that is a speedup of 3.75x over the collective network based algorithm in the product MPI stack on BG/Q.","PeriodicalId":243229,"journal":{"name":"European MPI Users Group Meeting","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121420981","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}