Proceedings of the 47th International Conference on Parallel Processing最新文献_第9页

Memory Coalescing for Hybrid Memory Cube 混合内存立方体的内存合并

Proceedings of the 47th International Conference on Parallel Processing Pub Date : 2018-08-13 DOI: 10.1145/3225058.3225062

Xi Wang, John D. Leidel, Yong Chen

{"title":"Memory Coalescing for Hybrid Memory Cube","authors":"Xi Wang, John D. Leidel, Yong Chen","doi":"10.1145/3225058.3225062","DOIUrl":"https://doi.org/10.1145/3225058.3225062","url":null,"abstract":"Arguably, many data-intensive applications pose significant challenges to conventional architectures and memory systems, especially when applications exhibit non-contiguous, irregular, and small memory access patterns. The long memory access latency can dramatically slow down the overall performance of applications. The growing desire of high memory bandwidth and low latency access stimulate the advent of novel 3D-staked memory devices such as the Hybrid Memory Cube (HMC), which provides significantly higher bandwidth compared with the conventional JEDEC DDR devices. Even though many existing studies have been devoted to achieving high bandwidth throughput of HMC, the bandwidth potential cannot be fully exploited due to the lack of highly efficient memory coalescing and interfacing methodology for HMC devices. In this research, we introduce a novel memory coalescer methodology that facilitates memory bandwidth efficiency and the overall performance through an efficient and scalable memory request coalescing interface for HMC. We present the design and implementation of this approach on RISC-V embedded cores with attached HMC devices. Our evaluation results show that the new memory coalescer eliminates 47.47% memory accesses to HMC and improves the overall performance by 13.14% on average.","PeriodicalId":307809,"journal":{"name":"Proceedings of the 47th International Conference on Parallel Processing","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125574774","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}

引用次数: 6

Interference between I/O and MPI Traffic on Fat-tree Networks 胖树网络中I/O和MPI流量之间的干扰

Proceedings of the 47th International Conference on Parallel Processing Pub Date : 2018-08-13 DOI: 10.1145/3225058.3225144

K. Brown, Nikhil Jain, S. Matsuoka, M. Schulz, A. Bhatele

引用次数: 6

Matrix Factorization on GPUs with Memory Optimization and Approximate Computing 基于内存优化和近似计算的gpu矩阵分解

Proceedings of the 47th International Conference on Parallel Processing Pub Date : 2018-08-11 DOI: 10.1145/3225058.3225096

Wei Tan, Shiyu Chang, L. Fong, Cheng Li, Zijun Wang, Liangliang Cao

{"title":"Matrix Factorization on GPUs with Memory Optimization and Approximate Computing","authors":"Wei Tan, Shiyu Chang, L. Fong, Cheng Li, Zijun Wang, Liangliang Cao","doi":"10.1145/3225058.3225096","DOIUrl":"https://doi.org/10.1145/3225058.3225096","url":null,"abstract":"Matrix factorization (MF) discovers latent features from observations, which has shown great promises in the fields of collaborative filtering, data compression, feature extraction, word embedding, etc. While many problem-specific optimization techniques have been proposed, alternating least square (ALS) remains popular due to its general applicability (e.g. easy to handle positive-unlabeled inputs), fast convergence and parallelization capability. Current MF implementations are either optimized for a single machine or with a need of a large computer cluster but still are insufficent. This is because a single machine provides limited compute power for large-scale data while multiple machines suffer from the network communication bottleneck. To address the aforementioned challenge, accelerating ALS on garphics processing units (GPUs) is a promising direction. We propose the novel approach in enhancing the MF efficiency via both memory optimization and approximate computing. The former exploits GPU memory hierarchy to increase data reuse, while the later reduces unneccessary computing without hurting the convergence of learning algorithms. Extensive experiments on large-scale datasets show that our solution not only outperforms the competing CPU solutions by a large margin but also has a 2x-4x performance gain compared to the state-of-the-art GPU solutions. Our implementations are open-sourced and publicly available.","PeriodicalId":307809,"journal":{"name":"Proceedings of the 47th International Conference on Parallel Processing","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-08-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114591898","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}

引用次数: 14

Balanced k-means for Parallel Geometric Partitioning 并行几何划分的平衡k-means

Proceedings of the 47th International Conference on Parallel Processing Pub Date : 2018-05-03 DOI: 10.1145/3225058.3225148

M. V. Looz, Charilaos Tzovas, Henning Meyerhenke

{"title":"Balanced k-means for Parallel Geometric Partitioning","authors":"M. V. Looz, Charilaos Tzovas, Henning Meyerhenke","doi":"10.1145/3225058.3225148","DOIUrl":"https://doi.org/10.1145/3225058.3225148","url":null,"abstract":"Mesh partitioning is an indispensable tool for efficient parallel numerical simulations. Its goal is to minimize communication between the processes of a simulation while achieving load balance. Established graph-based partitioning tools yield a high solution quality; however, their scalability is limited. Geometric approaches usually scale better, but their solution quality may be unsatisfactory for \"non-trivial\" mesh topologies. In this paper, we present a scalable version of k-means that is adapted to yield balanced clusters. Balanced k-means constitutes the core of our new partitioning algorithm Geographer. Bootstrapping of initial centers is performed with space-filling curves, leading to fast convergence of the subsequent balanced k-means algorithm. Our experiments with up to 16 384 MPI processes on numerous benchmark meshes show the following: (i) Geographer produces partitions with a lower communication volume than state-of-the-art geometric partitioners from the Zoltan package; (ii) Geographer scales well on large inputs; (iii) a Delaunay mesh with a few billion vertices and edges can be partitioned in a few seconds.","PeriodicalId":307809,"journal":{"name":"Proceedings of the 47th International Conference on Parallel Processing","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-05-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121082909","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}

引用次数: 8

Topology-induced Enhancement of Mappings 拓扑诱导的映射增强

Proceedings of the 47th International Conference on Parallel Processing Pub Date : 2018-04-19 DOI: 10.1145/3225058.3225117

R. Glantz, Maria Predari, Henning Meyerhenke

{"title":"Topology-induced Enhancement of Mappings","authors":"R. Glantz, Maria Predari, Henning Meyerhenke","doi":"10.1145/3225058.3225117","DOIUrl":"https://doi.org/10.1145/3225058.3225117","url":null,"abstract":"In this paper we propose a new method to enhance a mapping μ(·) of a parallel application's computational tasks to the processing elements (PEs) of a parallel computer. The idea behind our method TiMEr is to enhance such a mapping by drawing on the observation that many topologies take the form of a partial cube. This class of graphs includes all rectangular and cubic meshes, any such torus with even extensions in each dimension, all hypercubes, and all trees. Following previous work, we represent the parallel application and the parallel computer by graphs Ga and Gp, respectively. Gp being a partial cube allows us to label its vertices, the PEs, by bitvectors such that the cost of exchanging one unit of information between any two vertices of Gp amounts to the Hamming distance between their labels. By transferring these bitvectors to the vertex set of Ga via μ-1(·) and extending them to be unique on Ga, we can enhance μ(·) by swapping labels on Ga in a new way. Pairs of swapped labels are local w. r. t. the PEs, but not w. r. t. Ga. Moreover, permutations of the bitvectors' entries give rise to a plethora of hierarchies on the PEs. Through these hierarchies we turn TiMEr into a hierarchical method for improving μ(·) that is complementary to state-of-the-art methods for computing μ(·) in the first place. In our experiments we use TiMEr to enhance mappings of complex networks onto rectangular meshes and tori with 256 and 512 nodes, as well as hypercubes with 256 nodes. It turns out that common quality measure of mappings derived from state-of-the-art tools, such as Scotch and KaHIP, can be improved up to 34%.","PeriodicalId":307809,"journal":{"name":"Proceedings of the 47th International Conference on Parallel Processing","volume":"503 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129482228","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}

引用次数: 6

Implementing Push-Pull Efficiently in GraphBLAS 在GraphBLAS中高效实现推拉

Proceedings of the 47th International Conference on Parallel Processing Pub Date : 2018-04-10 DOI: 10.1145/3225058.3225122

Carl Yang, A. Buluç, John Douglas Owens

引用次数: 33

Task-parallel Analysis of Molecular Dynamics Trajectories 分子动力学轨迹的任务并行分析

Proceedings of the 47th International Conference on Parallel Processing Pub Date : 2018-01-23 DOI: 10.1145/3225058.3225128

Ioannis Paraskevakos, André Luckow, George Chantzialexiou, Mahzad Khoshlessan, O. Beckstein, G. Fox, S. Jha

{"title":"Task-parallel Analysis of Molecular Dynamics Trajectories","authors":"Ioannis Paraskevakos, André Luckow, George Chantzialexiou, Mahzad Khoshlessan, O. Beckstein, G. Fox, S. Jha","doi":"10.1145/3225058.3225128","DOIUrl":"https://doi.org/10.1145/3225058.3225128","url":null,"abstract":"Different parallel frameworks for implementing data analysis applications have been proposed by the HPC and Big Data communities. In this paper, we investigate three task-parallel frameworks: Spark, Dask and RADICAL-Pilot with respect to their ability to support data analytics on HPC resources and compare them to MPI. We investigate the data analysis requirements of Molecular Dynamics (MD) simulations which are significant consumers of supercomputing cycles, producing immense amounts of data. A typical large-scale MD simulation of a physical system of O(100k) atoms over μsecs can produce from O(10) GB to O(1000) GBs of data. We propose and evaluate different approaches for parallelization of a representative set of MD trajectory analysis algorithms, in particular the computation of path similarity and leaflet identification. We evaluate Spark, Dask and RADICAL-Pilot with respect to their abstractions and runtime engine capabilities to support these algorithms. We provide a conceptual basis for comparing and understanding different frameworks that enable users to select the optimal system for each application. We also provide a quantitative performance analysis of the different algorithms across the three frameworks.","PeriodicalId":307809,"journal":{"name":"Proceedings of the 47th International Conference on Parallel Processing","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-01-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123798124","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}

引用次数: 15

ImageNet Training in Minutes ImageNet培训分钟

Proceedings of the 47th International Conference on Parallel Processing Pub Date : 2017-09-14 DOI: 10.1145/3225058.3225069

Yang You, Zhao Zhang, Cho-Jui Hsieh, J. Demmel, K. Keutzer

引用次数: 401

IS-ASGD: Accelerating Asynchronous SGD using Importance Sampling IS-ASGD:使用重要性采样加速异步SGD

Proceedings of the 47th International Conference on Parallel Processing Pub Date : 2017-06-26 DOI: 10.1145/3225058.3225135

Fei Wang, Xiaofeng Gao, Jun Ye, Guihai Chen

{"title":"IS-ASGD: Accelerating Asynchronous SGD using Importance Sampling","authors":"Fei Wang, Xiaofeng Gao, Jun Ye, Guihai Chen","doi":"10.1145/3225058.3225135","DOIUrl":"https://doi.org/10.1145/3225058.3225135","url":null,"abstract":"Variance reduction (VR) techniques for convergence rate acceleration of stochastic gradient descent (SGD) algorithm have been developed with great efforts recently. VR's two variants, stochastic variance-reduced-gradient (SVRG-SGD) and importance sampling (IS-SGD) have achieved remarkable progresses. Meanwhile, asynchronous SGD (ASGD) is becoming more critical due to the ever-increasing scale of the optimization problems. The application of VR in ASGD to accelerate its convergence rate has therefore attracted much interest and SVRG-ASGDs were proposed. However, we found that SVRG suffers dissatisfying performance in accelerating ASGD when datasets are sparse and large-scale. In such case, SVRG-ASGD's iterative computation cost is magnitudes higher than plain ASGD which makes it very inefficient. On the other hand, IS achieves improved convergence rate with few extra computation cost and is invariant to the sparsity of datasets. These advantages make it very suitable for the acceleration of ASGD on large-scale sparse datasets. In this paper we propose a novel IS-combined ASGD for efficient convergence rate acceleration, namely, IS-ASGD. We theoretically prove the superior convergence bound of IS-ASGD. Experimental results also demonstrate our statements.","PeriodicalId":307809,"journal":{"name":"Proceedings of the 47th International Conference on Parallel Processing","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130790080","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}

引用次数: 4