首页 > 最新文献
paperpass

2012 International Conference for High Performance Computing, Networking, Storage and Analysis最新文献

英文 中文
Byte-precision level of detail processing for variable precision analytics 用于可变精度分析的字节精度级别的细节处理
John Jenkins, Eric R. Schendel, Sriram Lakshminarasimhan, David A. Boyuka, Terry Rogers, S. Ethier, R. Ross, S. Klasky, N. Samatova
I/O bottlenecks in HPC applications are becoming a more pressing problem as compute capabilities continue to outpace I/O capabilities. While double-precision simulation data often must be stored losslessly, the loss of some of the fractional component may introduce acceptably small errors to many types of scientific analyses. Given this observation, we develop a precision level of detail (APLOD) library, which partitions double-precision datasets along user-defined byte boundaries. APLOD parameterizes the analysis accuracy-I/O performance tradeoff, bounds maximum relative error, maintains I/O access patterns compared to full precision, and operates with low overhead. Using ADIOS as an I/O use-case, we show proportional reduction in disk access time to the degree of precision. Finally, we show the effects of partial precision analysis on accuracy for operations such as k-means and Fourier analysis, finding a strong applicability for the use of varying degrees of precision to reduce the cost of analyzing extreme-scale data.
随着计算能力继续超过I/O能力,HPC应用程序中的I/O瓶颈正在成为一个更加紧迫的问题。虽然双精度模拟数据通常必须无损存储,但一些分数分量的损失可能会给许多类型的科学分析带来可接受的小误差。考虑到这一点,我们开发了一个精度级别的细节(APLOD)库,它沿着用户定义的字节边界划分双精度数据集。APLOD参数化了分析精度与I/O性能之间的权衡,限制了最大相对误差,维护了与完全精度相比的I/O访问模式,并以较低的开销运行。使用ADIOS作为I/O用例,我们展示了磁盘访问时间按比例减少的精度程度。最后,我们展示了部分精度分析对k-means和傅立叶分析等操作精度的影响,发现使用不同精度程度来降低分析极端尺度数据的成本具有很强的适用性。
{"title":"Byte-precision level of detail processing for variable precision analytics","authors":"John Jenkins, Eric R. Schendel, Sriram Lakshminarasimhan, David A. Boyuka, Terry Rogers, S. Ethier, R. Ross, S. Klasky, N. Samatova","doi":"10.1109/SC.2012.26","DOIUrl":"https://doi.org/10.1109/SC.2012.26","url":null,"abstract":"I/O bottlenecks in HPC applications are becoming a more pressing problem as compute capabilities continue to outpace I/O capabilities. While double-precision simulation data often must be stored losslessly, the loss of some of the fractional component may introduce acceptably small errors to many types of scientific analyses. Given this observation, we develop a precision level of detail (APLOD) library, which partitions double-precision datasets along user-defined byte boundaries. APLOD parameterizes the analysis accuracy-I/O performance tradeoff, bounds maximum relative error, maintains I/O access patterns compared to full precision, and operates with low overhead. Using ADIOS as an I/O use-case, we show proportional reduction in disk access time to the degree of precision. Finally, we show the effects of partial precision analysis on accuracy for operations such as k-means and Fourier analysis, finding a strong applicability for the use of varying degrees of precision to reduce the cost of analyzing extreme-scale data.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124000424","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: 29
Hybridizing S3D into an Exascale application using OpenACC: An approach for moving to multi-petaflops and beyond 使用OpenACC将S3D混合到Exascale应用程序中:一种移动到千万亿次浮点运算甚至更高的方法
J. Levesque, R. Sankaran, R. Grout
Hybridization is the process of converting an application with a single level of parallelism to an application with multiple levels of parallelism. Over the past 15 years a majority of the applications that run on High Performance Computing systems have employed MPI for all of the parallelism within the application. In the Peta-Exascale computing regime, effective utilization of the hardware requires multiple levels of parallelism matched to the macro architecture of the system to achieve good performance. A hybridized code base is performance portable when sufficient parallelism is expressed in an architecture agnostic form to achieve good performance on a range of available systems. The hybridized S3D code is performance portable across today's leading many core and GPU accelerated systems. The OpenACC framework allows a unified code base to be deployed for either (Manycore CPU or Manycore CPU+GPU) while permitting architecture specific optimizations to expose new dimensions of parallelism to be utilized.
杂交是将具有单一并行级别的应用程序转换为具有多个并行级别的应用程序的过程。在过去的15年中,大多数运行在高性能计算系统上的应用程序都使用MPI来实现应用程序中的所有并行性。在Peta-Exascale计算机制中,有效利用硬件需要与系统的宏观体系结构相匹配的多级并行性才能获得良好的性能。当以与体系结构无关的形式表达足够的并行性以在一系列可用系统上实现良好的性能时,混合代码库具有性能可移植性。混合的S3D代码可在当今领先的许多核心和GPU加速系统上进行性能移植。OpenACC框架允许为(多核CPU或多核CPU+GPU)部署统一的代码库,同时允许架构特定的优化来暴露新的并行性维度。
{"title":"Hybridizing S3D into an Exascale application using OpenACC: An approach for moving to multi-petaflops and beyond","authors":"J. Levesque, R. Sankaran, R. Grout","doi":"10.1109/SC.2012.69","DOIUrl":"https://doi.org/10.1109/SC.2012.69","url":null,"abstract":"Hybridization is the process of converting an application with a single level of parallelism to an application with multiple levels of parallelism. Over the past 15 years a majority of the applications that run on High Performance Computing systems have employed MPI for all of the parallelism within the application. In the Peta-Exascale computing regime, effective utilization of the hardware requires multiple levels of parallelism matched to the macro architecture of the system to achieve good performance. A hybridized code base is performance portable when sufficient parallelism is expressed in an architecture agnostic form to achieve good performance on a range of available systems. The hybridized S3D code is performance portable across today's leading many core and GPU accelerated systems. The OpenACC framework allows a unified code base to be deployed for either (Manycore CPU or Manycore CPU+GPU) while permitting architecture specific optimizations to expose new dimensions of parallelism to be utilized.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123089001","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: 61
On using virtual circuits for GridFTP transfers 关于在GridFTP传输中使用虚拟电路
Zhengyang Liu, M. Veeraraghavan, Zhenzhen Yan, Chris Tracy, J. Tie, Ian T Foster, J. Dennis, Jason Hick, Yee-Ting Li, W. Yang
The goal of this work is to characterize scientific data transfers and to determine the suitability of dynamic virtual circuit service for these transfers instead of the currently used IP-routed service. Specifically, logs collected by servers executing a commonly used scientific data transfer application, GridFTP, are obtained from three US super-computing/scientific research centers, NERSC, SLAC, and NCAR, and analyzed. Dynamic virtual circuit (VC) service, a relatively new offering from providers such as ESnet and Internet2, allows for the selection of a path on which a rate-guaranteed connection is established prior to data transfer. Given VC setup overhead, the first analysis of the GridFTP transfer logs characterizes the duration of sessions, where a session consists of multiple back-to-back transfers executed in batch mode between the same two GridFTP servers. Of the NCAR-NICS sessions analyzed, 56% of all sessions (90% of all transfers) would have been long enough to be served with dynamic VC service. An analysis of transfer logs across four paths, NCAR-NICS, SLAC-BNL, NERSC-ORNL and NERSC-ANL, shows significant throughput variance, where NICS, BNL, ORNL, and ANL are other US national laboratories. For example, on the NERSC-ORNL path, the inter-quartile range was 695 Mbps, with a maximum value of 3.64 Gbps and a minimum value of 758 Mbps. An analysis of the impact of various factors that are potential causes of this variance is also presented.
这项工作的目标是表征科学数据传输,并确定动态虚拟电路服务对这些传输的适用性,而不是目前使用的ip路由服务。具体来说,执行常用科学数据传输应用程序GridFTP的服务器收集的日志是从美国三个超级计算/科学研究中心(NERSC、SLAC和NCAR)获取并分析的。动态虚拟电路(VC)服务是ESnet和Internet2等提供商提供的一种相对较新的服务,它允许选择在数据传输之前建立速率保证连接的路径。考虑到VC设置开销,对GridFTP传输日志的第一个分析描述了会话的持续时间,其中会话由在相同的两个GridFTP服务器之间以批处理模式执行的多个背对背传输组成。在分析的NCAR-NICS会话中,56%的会话(90%的传输)足够长,可以使用动态VC服务。对四个路径(NCAR-NICS、SLAC-BNL、NERSC-ORNL和NERSC-ANL)的传输日志进行分析,显示出显著的吞吐量差异,其中NICS、BNL、ORNL和ANL是其他美国国家实验室。例如,在NERSC-ORNL路径上,四分位数间距为695 Mbps,最大值为3.64 Gbps,最小值为758 Mbps。分析了各种因素的影响,这些因素是造成这种差异的潜在原因。
{"title":"On using virtual circuits for GridFTP transfers","authors":"Zhengyang Liu, M. Veeraraghavan, Zhenzhen Yan, Chris Tracy, J. Tie, Ian T Foster, J. Dennis, Jason Hick, Yee-Ting Li, W. Yang","doi":"10.5555/2388996.2389106","DOIUrl":"https://doi.org/10.5555/2388996.2389106","url":null,"abstract":"The goal of this work is to characterize scientific data transfers and to determine the suitability of dynamic virtual circuit service for these transfers instead of the currently used IP-routed service. Specifically, logs collected by servers executing a commonly used scientific data transfer application, GridFTP, are obtained from three US super-computing/scientific research centers, NERSC, SLAC, and NCAR, and analyzed. Dynamic virtual circuit (VC) service, a relatively new offering from providers such as ESnet and Internet2, allows for the selection of a path on which a rate-guaranteed connection is established prior to data transfer. Given VC setup overhead, the first analysis of the GridFTP transfer logs characterizes the duration of sessions, where a session consists of multiple back-to-back transfers executed in batch mode between the same two GridFTP servers. Of the NCAR-NICS sessions analyzed, 56% of all sessions (90% of all transfers) would have been long enough to be served with dynamic VC service. An analysis of transfer logs across four paths, NCAR-NICS, SLAC-BNL, NERSC-ORNL and NERSC-ANL, shows significant throughput variance, where NICS, BNL, ORNL, and ANL are other US national laboratories. For example, on the NERSC-ORNL path, the inter-quartile range was 695 Mbps, with a maximum value of 3.64 Gbps and a minimum value of 758 Mbps. An analysis of the impact of various factors that are potential causes of this variance is also presented.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"118 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128117941","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: 13
RAMZzz: Rank-aware DRAM power management with dynamic migrations and demotions 具有动态迁移和降级的等级感知DRAM电源管理
Donghong Wu, Bingsheng He, Xueyan Tang, Jianliang Xu, M. Guo
Main memory is a significant energy consumer which may contribute to over 40% of the total system power, and will become more significant for server machines with more main memory. In this paper, we propose a novel memory system design named RAMZzz with rank-aware energy saving optimizations. Specifically, we rely on a memory controller to monitor the memory access locality, and group the pages with similar access locality into the same rank. We further develop dynamic page migrations to adapt to data access patterns, and a prediction model to estimate the demotion time for accurate control on power state transitions. We experimentally compare our algorithm with other energy saving policies with cycle-accurate simulation. Experiments with benchmark workloads show that RAMZzz achieves significant improvement on energy-delay2 and energy consumption over other power saving techniques.
主存是一个重要的能源消耗者,它可能贡献了超过40%的系统总功率,并且对于具有更多主存的服务器机器将变得更加重要。在本文中,我们提出了一种新的存储系统设计,称为RAMZzz,具有等级感知的节能优化。具体来说,我们依靠内存控制器来监视内存访问位置,并将具有相似访问位置的页面分组到相同的级别。我们进一步开发了动态页面迁移以适应数据访问模式,并开发了一个预测模型来估计降级时间,以精确控制电源状态转换。通过循环精确仿真,将该算法与其他节能策略进行了实验比较。对基准工作负载的实验表明,RAMZzz比其他节能技术在能源延迟2和能源消耗方面取得了显著的改进。
{"title":"RAMZzz: Rank-aware DRAM power management with dynamic migrations and demotions","authors":"Donghong Wu, Bingsheng He, Xueyan Tang, Jianliang Xu, M. Guo","doi":"10.1109/SC.2012.99","DOIUrl":"https://doi.org/10.1109/SC.2012.99","url":null,"abstract":"Main memory is a significant energy consumer which may contribute to over 40% of the total system power, and will become more significant for server machines with more main memory. In this paper, we propose a novel memory system design named RAMZzz with rank-aware energy saving optimizations. Specifically, we rely on a memory controller to monitor the memory access locality, and group the pages with similar access locality into the same rank. We further develop dynamic page migrations to adapt to data access patterns, and a prediction model to estimate the demotion time for accurate control on power state transitions. We experimentally compare our algorithm with other energy saving policies with cycle-accurate simulation. Experiments with benchmark workloads show that RAMZzz achieves significant improvement on energy-delay2 and energy consumption over other power saving techniques.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130632466","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: 30
Managing data-movement for effective shared-memory parallelization of out-of-core sparse solvers 管理数据移动,实现核外稀疏求解器的有效共享内存并行化
H. Avron, Anshul Gupta
Direct methods for solving sparse linear systems are robust and typically exhibit good performance, but often require large amounts of memory due to fill-in. Many industrial applications use out-of-core techniques to mitigate this problem. However, parallelizing sparse out-of-core solvers poses some unique challenges because accessing secondary storage introduces serialization and I/O overhead. We analyze the data-movement costs and memory versus parallelism trade-offs in a shared-memory parallel out-of-core linear solver for sparse symmetric systems. We propose an algorithm that uses a novel memory management scheme and adaptive task parallelism to reduce the data-movement costs. We present experiments to show that our solver is faster than existing out-of-core sparse solvers on a single core, and is more scalable than the only other known shared-memory parallel out-of-core solver. This work is also directly applicable at the node level in a distributed-memory parallel scenario.
求解稀疏线性系统的直接方法是鲁棒的,通常表现出良好的性能,但由于填充通常需要大量的内存。许多工业应用程序使用外核技术来缓解这个问题。然而,并行化稀疏的out-of-core解决方案会带来一些独特的挑战,因为访问二级存储会引入序列化和I/O开销。我们分析了稀疏对称系统的共享内存并行外核线性求解器的数据移动成本和内存与并行性的权衡。我们提出了一种使用新的内存管理方案和自适应任务并行性来降低数据移动成本的算法。我们提出的实验表明,我们的求解器在单核上比现有的out- core稀疏求解器更快,并且比唯一已知的共享内存并行out- core求解器更具可扩展性。这项工作也直接适用于分布式内存并行场景中的节点级别。
{"title":"Managing data-movement for effective shared-memory parallelization of out-of-core sparse solvers","authors":"H. Avron, Anshul Gupta","doi":"10.1109/SC.2012.74","DOIUrl":"https://doi.org/10.1109/SC.2012.74","url":null,"abstract":"Direct methods for solving sparse linear systems are robust and typically exhibit good performance, but often require large amounts of memory due to fill-in. Many industrial applications use out-of-core techniques to mitigate this problem. However, parallelizing sparse out-of-core solvers poses some unique challenges because accessing secondary storage introduces serialization and I/O overhead. We analyze the data-movement costs and memory versus parallelism trade-offs in a shared-memory parallel out-of-core linear solver for sparse symmetric systems. We propose an algorithm that uses a novel memory management scheme and adaptive task parallelism to reduce the data-movement costs. We present experiments to show that our solver is faster than existing out-of-core sparse solvers on a single core, and is more scalable than the only other known shared-memory parallel out-of-core solver. This work is also directly applicable at the node level in a distributed-memory parallel scenario.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116971038","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: 11
Communication-Avoiding Parallel Strassen: Implementation and performance 避免通信并行Strassen:实现和性能
Benjamin Lipshitz, Grey Ballard, J. Demmel, O. Schwartz
Matrix multiplication is a fundamental kernel of many high performance and scientific computing applications. Most parallel implementations use classical O(n3) matrix multiplication, even though there exist algorithms with lower arithmetic complexity. We recently presented a new Communication-Avoiding Parallel Strassen algorithm (CAPS), based on Strassen's fast matrix multiplication, that minimizes communication (SPAA'12). It communicates asymptotically less than all classical and all previous Strassen-based algorithms, and it attains theoretical lower bounds. In this paper we show that CAPS is also faster in practice. We benchmark and compare its performance to previous algorithms on Hopper (Cray XE6), Intrepid (IBM BG/P), and Franklin (Cray XT4). We demonstrate significant speedups over previous algorithms both for large matrices and for small matrices on large numbers of processors. We model and analyze the performance of CAPS and predict its performance on future exascale platforms.
矩阵乘法是许多高性能和科学计算应用的基本核心。大多数并行实现使用经典的O(n3)矩阵乘法,即使存在算术复杂度较低的算法。我们最近提出了一种新的通信避免并行Strassen算法(CAPS),基于Strassen的快速矩阵乘法,最小化通信(SPAA'12)。它的通信渐近性小于所有经典算法和所有先前的基于strassen的算法,并且它达到了理论下界。在本文中,我们证明了CAPS在实践中也更快。我们将其性能与Hopper (Cray XE6)、Intrepid (IBM BG/P)和Franklin (Cray XT4)上的先前算法进行基准测试和比较。我们展示了在大量处理器上的大矩阵和小矩阵比以前的算法有显著的加速。我们对CAPS的性能进行建模和分析,并预测其在未来百亿亿级平台上的性能。
{"title":"Communication-Avoiding Parallel Strassen: Implementation and performance","authors":"Benjamin Lipshitz, Grey Ballard, J. Demmel, O. Schwartz","doi":"10.1109/SC.2012.33","DOIUrl":"https://doi.org/10.1109/SC.2012.33","url":null,"abstract":"Matrix multiplication is a fundamental kernel of many high performance and scientific computing applications. Most parallel implementations use classical O(n3) matrix multiplication, even though there exist algorithms with lower arithmetic complexity. We recently presented a new Communication-Avoiding Parallel Strassen algorithm (CAPS), based on Strassen's fast matrix multiplication, that minimizes communication (SPAA'12). It communicates asymptotically less than all classical and all previous Strassen-based algorithms, and it attains theoretical lower bounds. In this paper we show that CAPS is also faster in practice. We benchmark and compare its performance to previous algorithms on Hopper (Cray XE6), Intrepid (IBM BG/P), and Franklin (Cray XT4). We demonstrate significant speedups over previous algorithms both for large matrices and for small matrices on large numbers of processors. We model and analyze the performance of CAPS and predict its performance on future exascale platforms.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127744879","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: 54
A massively space-time parallel N-body solver 一个大规模时空平行n体求解器
R. Speck, D. Ruprecht, R. Krause, M. Emmett, M. Minion, M. Winkel, P. Gibbon
We present a novel space-time parallel version of the Barnes-Hut tree code PEPC using PFASST, the Parallel Full Approximation Scheme in Space and Time. The naive use of increasingly more processors for a fixed-size N-body problem is prone to saturate as soon as the number of unknowns per core becomes too small. To overcome this intrinsic strong-scaling limit, we introduce temporal parallelism on top of PEPC's existing hybrid MPI/PThreads spatial decomposition. Here, we use PFASST which is based on a combination of the iterations of the parallel-in-time algorithm parareal with the sweeps of spectral deferred correction (SDC) schemes. By combining these sweeps with multiple space-time discretization levels, PFASST relaxes the theoretical bound on parallel efficiency in parareal. We present results from runs on up to 262,144 cores on the IBM Blue Gene/P installation JUGENE, demonstrating that the spacetime parallel code provides speedup beyond the saturation of the purely space-parallel approach.
我们提出了一种新的时空并行版本的Barnes-Hut树代码PEPC,使用pfast,空间和时间的并行全近似格式。对于固定大小的n体问题,天真地使用越来越多的处理器,一旦每个核心的未知数数量变得太小,就容易达到饱和。为了克服这种固有的强尺度限制,我们在PEPC现有的混合MPI/PThreads空间分解之上引入了时间并行性。在这里,我们使用PFASST,它是基于并行实时算法的迭代与光谱延迟校正(SDC)方案的扫描的平行组合。通过将这些扫描与多个时空离散化水平相结合,pfast放宽了平行效率的理论界限。我们展示了在IBM Blue Gene/P安装JUGENE上多达262,144个内核上运行的结果,证明了时空并行代码提供了超越纯空间并行方法饱和的加速。
{"title":"A massively space-time parallel N-body solver","authors":"R. Speck, D. Ruprecht, R. Krause, M. Emmett, M. Minion, M. Winkel, P. Gibbon","doi":"10.1109/SC.2012.6","DOIUrl":"https://doi.org/10.1109/SC.2012.6","url":null,"abstract":"We present a novel space-time parallel version of the Barnes-Hut tree code PEPC using PFASST, the Parallel Full Approximation Scheme in Space and Time. The naive use of increasingly more processors for a fixed-size N-body problem is prone to saturate as soon as the number of unknowns per core becomes too small. To overcome this intrinsic strong-scaling limit, we introduce temporal parallelism on top of PEPC's existing hybrid MPI/PThreads spatial decomposition. Here, we use PFASST which is based on a combination of the iterations of the parallel-in-time algorithm parareal with the sweeps of spectral deferred correction (SDC) schemes. By combining these sweeps with multiple space-time discretization levels, PFASST relaxes the theoretical bound on parallel efficiency in parareal. We present results from runs on up to 262,144 cores on the IBM Blue Gene/P installation JUGENE, demonstrating that the spacetime parallel code provides speedup beyond the saturation of the purely space-parallel approach.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129560537","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: 76
Tiling stencil computations to maximize parallelism 平铺模板计算以最大化并行性
Vinayaka Bandishti, Irshad Pananilath, Uday Bondhugula
Most stencil computations allow tile-wise concurrent start, i.e., there always exists a face of the iteration space and a set of tiling hyperplanes such that all tiles along that face can be started concurrently. This provides load balance and maximizes parallelism. However, existing automatic tiling frameworks often choose hyperplanes that lead to pipelined start-up and load imbalance. We address this issue with a new tiling technique that ensures concurrent start-up as well as perfect load-balance whenever possible. We first provide necessary and sufficient conditions on tiling hyperplanes to enable concurrent start for programs with affine data accesses. We then provide an approach to find such hyperplanes. Experimental evaluation on a 12-core Intel Westmere shows that our code is able to outperform a tuned domain-specific stencil code generator by 4% to 27%, and previous compiler techniques by a factor of 2× to 10.14×.
大多数模板计算允许逐块并发启动,也就是说,总是存在迭代空间的一个面和一组平铺超平面,这样沿着该面的所有平铺都可以并发启动。这提供了负载平衡和最大化并行性。然而,现有的自动平铺框架经常选择导致流水线启动和负载不平衡的超平面。我们用一种新的平铺技术来解决这个问题,这种技术可以确保并发启动以及尽可能完美的负载平衡。我们首先提供了平铺超平面的充分必要条件,使具有仿射数据访问的程序能够并发启动。然后我们提供了一种方法来找到这样的超平面。在12核Intel Westmere上的实验评估表明,我们的代码能够比经过调优的特定领域模板代码生成器的性能高出4%到27%,并且比以前的编译器技术高出2倍到10.14倍。
{"title":"Tiling stencil computations to maximize parallelism","authors":"Vinayaka Bandishti, Irshad Pananilath, Uday Bondhugula","doi":"10.1109/SC.2012.107","DOIUrl":"https://doi.org/10.1109/SC.2012.107","url":null,"abstract":"Most stencil computations allow tile-wise concurrent start, i.e., there always exists a face of the iteration space and a set of tiling hyperplanes such that all tiles along that face can be started concurrently. This provides load balance and maximizes parallelism. However, existing automatic tiling frameworks often choose hyperplanes that lead to pipelined start-up and load imbalance. We address this issue with a new tiling technique that ensures concurrent start-up as well as perfect load-balance whenever possible. We first provide necessary and sufficient conditions on tiling hyperplanes to enable concurrent start for programs with affine data accesses. We then provide an approach to find such hyperplanes. Experimental evaluation on a 12-core Intel Westmere shows that our code is able to outperform a tuned domain-specific stencil code generator by 4% to 27%, and previous compiler techniques by a factor of 2× to 10.14×.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125876458","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: 142
A new scalable parallel DBSCAN algorithm using the disjoint-set data structure 一种新的基于不相交集数据结构的可扩展并行DBSCAN算法
Md. Mostofa Ali Patwary, Diana Palsetia, Ankit Agrawal, W. Liao, F. Manne, A. Choudhary
DBSCAN is a well-known density based clustering algorithm capable of discovering arbitrary shaped clusters and eliminating noise data. However, parallelization of DBSCAN is challenging as it exhibits an inherent sequential data access order. Moreover, existing parallel implementations adopt a master-slave strategy which can easily cause an unbalanced workload and hence result in low parallel efficiency. We present a new parallel DBSCAN algorithm (PDSDBSCAN) using graph algorithmic concepts. More specifically, we employ the disjoint-set data structure to break the access sequentiality of DBSCAN. In addition, we use a tree-based bottom-up approach to construct the clusters. This yields a better-balanced workload distribution. We implement the algorithm both for shared and for distributed memory. Using data sets containing up to several hundred million high-dimensional points, we show that PDSDBSCAN significantly outperforms the master-slave approach, achieving speedups up to 25.97 using 40 cores on shared memory architecture, and speedups up to 5,765 using 8,192 cores on distributed memory architecture.
DBSCAN是一种著名的基于密度的聚类算法,能够发现任意形状的聚类并消除噪声数据。但是,DBSCAN的并行化具有挑战性,因为它显示了固有的顺序数据访问顺序。此外,现有的并行实现采用主从策略,这很容易导致工作负载不平衡,从而导致并行效率低下。利用图算法的概念提出了一种新的并行DBSCAN算法(PDSDBSCAN)。更具体地说,我们采用了不相交集的数据结构来打破DBSCAN的访问顺序。此外,我们使用基于树的自底向上方法来构建集群。这将产生更均衡的工作负载分布。我们对共享内存和分布式内存都实现了该算法。使用包含多达数亿个高维点的数据集,我们表明PDSDBSCAN显著优于主从方法,在共享内存架构上使用40个内核实现高达25.97的加速,在分布式内存架构上使用8,192个内核实现高达5,765的加速。
{"title":"A new scalable parallel DBSCAN algorithm using the disjoint-set data structure","authors":"Md. Mostofa Ali Patwary, Diana Palsetia, Ankit Agrawal, W. Liao, F. Manne, A. Choudhary","doi":"10.1109/SC.2012.9","DOIUrl":"https://doi.org/10.1109/SC.2012.9","url":null,"abstract":"DBSCAN is a well-known density based clustering algorithm capable of discovering arbitrary shaped clusters and eliminating noise data. However, parallelization of DBSCAN is challenging as it exhibits an inherent sequential data access order. Moreover, existing parallel implementations adopt a master-slave strategy which can easily cause an unbalanced workload and hence result in low parallel efficiency. We present a new parallel DBSCAN algorithm (PDSDBSCAN) using graph algorithmic concepts. More specifically, we employ the disjoint-set data structure to break the access sequentiality of DBSCAN. In addition, we use a tree-based bottom-up approach to construct the clusters. This yields a better-balanced workload distribution. We implement the algorithm both for shared and for distributed memory. Using data sets containing up to several hundred million high-dimensional points, we show that PDSDBSCAN significantly outperforms the master-slave approach, achieving speedups up to 25.97 using 40 cores on shared memory architecture, and speedups up to 5,765 using 8,192 cores on distributed memory architecture.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121737701","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: 168
MAGE: Adaptive Granularity and ECC for resilient and power efficient memory systems MAGE:自适应粒度和ECC用于弹性和节能内存系统
Sheng Li, D. Yoon, Ke Chen, Jishen Zhao, Jung Ho Ahn, J. Brockman, Yuan Xie, N. Jouppi
Resiliency is one of the toughest challenges in high-performance computing, and memory accounts for a significant fraction of errors. Providing strong error tolerance in memory usually requires a wide memory channel that incurs a large access granularity (hence, a large cache line). Unfortunately, applications with limited spatial locality waste memory power and bandwidth on systems with a large access granularity. Thus, careful design considerations must be made to balance memory system performance, power efficiency, and resiliency. In this paper, we propose MAGE, a Memory system with Adaptive Granularity and ECC, to achieve high performance, power efficiency, and resiliency. MAGE can adapt memory access granularities and ECC schemes to applications with different memory behaviors. Our experiments show that MAGE achieves more than a 28% energy-delay product improvement, compared to the best existing systems with static granularity and ECC.
弹性是高性能计算中最棘手的挑战之一,内存占错误的很大一部分。在内存中提供强大的容错性通常需要宽的内存通道,这会导致大的访问粒度(因此需要大的缓存线)。不幸的是,具有有限空间局部性的应用程序会在具有大访问粒度的系统上浪费内存和带宽。因此,必须仔细考虑设计,以平衡内存系统性能、电源效率和弹性。在本文中,我们提出了MAGE,一个具有自适应粒度和ECC的内存系统,以实现高性能,功耗效率和弹性。MAGE可以根据不同的内存行为调整内存访问粒度和ECC方案。我们的实验表明,与具有静态粒度和ECC的现有最佳系统相比,MAGE实现了28%以上的能量延迟产品改进。
{"title":"MAGE: Adaptive Granularity and ECC for resilient and power efficient memory systems","authors":"Sheng Li, D. Yoon, Ke Chen, Jishen Zhao, Jung Ho Ahn, J. Brockman, Yuan Xie, N. Jouppi","doi":"10.1109/SC.2012.73","DOIUrl":"https://doi.org/10.1109/SC.2012.73","url":null,"abstract":"Resiliency is one of the toughest challenges in high-performance computing, and memory accounts for a significant fraction of errors. Providing strong error tolerance in memory usually requires a wide memory channel that incurs a large access granularity (hence, a large cache line). Unfortunately, applications with limited spatial locality waste memory power and bandwidth on systems with a large access granularity. Thus, careful design considerations must be made to balance memory system performance, power efficiency, and resiliency. In this paper, we propose MAGE, a Memory system with Adaptive Granularity and ECC, to achieve high performance, power efficiency, and resiliency. MAGE can adapt memory access granularities and ECC schemes to applications with different memory behaviors. Our experiments show that MAGE achieves more than a 28% energy-delay product improvement, compared to the best existing systems with static granularity and ECC.","PeriodicalId":217221,"journal":{"name":"2012 International Conference for High Performance Computing, Networking, Storage and Analysis","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122212210","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: 28
期刊
2012 International Conference for High Performance Computing, Networking, Storage and Analysis
全部 Acc. Chem. Res. ACS Applied Bio Materials ACS Appl. Electron. Mater. ACS Appl. Energy Mater. ACS Appl. Mater. Interfaces ACS Appl. Nano Mater. ACS Appl. Polym. Mater. ACS BIOMATER-SCI ENG ACS Catal. ACS Cent. Sci. ACS Chem. Biol. ACS Chemical Health & Safety ACS Chem. Neurosci. ACS Comb. Sci. ACS Earth Space Chem. ACS Energy Lett. ACS Infect. Dis. ACS Macro Lett. ACS Mater. Lett. ACS Med. Chem. Lett. ACS Nano ACS Omega ACS Photonics ACS Sens. ACS Sustainable Chem. Eng. ACS Synth. Biol. Anal. Chem. BIOCHEMISTRY-US Bioconjugate Chem. BIOMACROMOLECULES Chem. Res. Toxicol. Chem. Rev. Chem. Mater. CRYST GROWTH DES ENERG FUEL Environ. Sci. Technol. Environ. Sci. Technol. Lett. Eur. J. Inorg. Chem. IND ENG CHEM RES Inorg. Chem. J. Agric. Food. Chem. J. Chem. Eng. Data J. Chem. Educ. J. Chem. Inf. Model. J. Chem. Theory Comput. J. Med. Chem. J. Nat. Prod. J PROTEOME RES J. Am. Chem. Soc. LANGMUIR MACROMOLECULES Mol. Pharmaceutics Nano Lett. Org. Lett. ORG PROCESS RES DEV ORGANOMETALLICS J. Org. Chem. J. Phys. Chem. J. Phys. Chem. A J. Phys. Chem. B J. Phys. Chem. C J. Phys. Chem. Lett. Analyst Anal. Methods Biomater. Sci. Catal. Sci. Technol. Chem. Commun. Chem. Soc. Rev. CHEM EDUC RES PRACT CRYSTENGCOMM Dalton Trans. Energy Environ. Sci. ENVIRON SCI-NANO ENVIRON SCI-PROC IMP ENVIRON SCI-WAT RES Faraday Discuss. Food Funct. Green Chem. Inorg. Chem. Front. Integr. Biol. J. Anal. At. Spectrom. J. Mater. Chem. A J. Mater. Chem. B J. Mater. Chem. C Lab Chip Mater. Chem. Front. Mater. Horiz. MEDCHEMCOMM Metallomics Mol. Biosyst. Mol. Syst. Des. Eng. Nanoscale Nanoscale Horiz. Nat. Prod. Rep. New J. Chem. Org. Biomol. Chem. Org. Chem. Front. PHOTOCH PHOTOBIO SCI PCCP Polym. Chem.
﹀
mjeditor
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
Book学术官方微信
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1