Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916270
A. Abdelfattah, S. Tomov, J. Dongarra
This paper presents a progressive approach for optimizing the batched LU factorization on graphics processing units (GPUs). The paper shows that the reliance on level-3 BLAS routines for performance does not really pay off, and that it is indeed important to pay attention to the memory-bound part of the algorithm, especially when the problem size is very small. In this context, we develop a size-aware multi-level blocking technique that utilizes different granularities for kernel fusion according to the problem size. Our experiments, which are conducted on a Tesla V100 GPU, show that the multi-level blocking technique achieves speedups for single/double precisions that are up to 3.28×/2.69× against the generic LAPACK-style implementation. It is also up to 8.72×/7.2× faster than the cuBLAS library for single and double precisions, respectively. The developed solution is integrated into the open-source MAGMA library.
{"title":"Progressive Optimization of Batched LU Factorization on GPUs","authors":"A. Abdelfattah, S. Tomov, J. Dongarra","doi":"10.1109/HPEC.2019.8916270","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916270","url":null,"abstract":"This paper presents a progressive approach for optimizing the batched LU factorization on graphics processing units (GPUs). The paper shows that the reliance on level-3 BLAS routines for performance does not really pay off, and that it is indeed important to pay attention to the memory-bound part of the algorithm, especially when the problem size is very small. In this context, we develop a size-aware multi-level blocking technique that utilizes different granularities for kernel fusion according to the problem size. Our experiments, which are conducted on a Tesla V100 GPU, show that the multi-level blocking technique achieves speedups for single/double precisions that are up to 3.28×/2.69× against the generic LAPACK-style implementation. It is also up to 8.72×/7.2× faster than the cuBLAS library for single and double precisions, respectively. The developed solution is integrated into the open-source MAGMA library.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"1996 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127322379","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}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916467
V. Kotteda, Anitha Kommu, Vinod Kumar
We carried out a nondeterministic analysis of flow in a fluidized bed. The flow in the fluidized bed is simulated with National Energy Technology Laboratory’s open-source multiphase fluid dynamics suite MFiX. It does not possess tools for uncertainty quantification. Therefore, we developed a C++ wrapper to integrate an uncertainty quantification toolkit developed at Sandia National Laboratory with MFiX. The wrapper exchanges uncertain input parameters and key output parameters among Dakota and MFiX. However, a data-driven framework is also developed to obtain reliable statistics as it is not feasible to get them with MFiX integrated into Dakota, Dakota-MFiX. The data generated from Dakota-MFiX simulations, with the Latin Hypercube method of sampling size 500, is used to train a machine-learning algorithm. The trained and tested deep neural network algorithm is integrated with Dakota via the wrapper to obtain low order statistics of the bed height and pressure drop across the bed.
{"title":"A data-driven framework for uncertainty quantification of a fluidized bed","authors":"V. Kotteda, Anitha Kommu, Vinod Kumar","doi":"10.1109/HPEC.2019.8916467","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916467","url":null,"abstract":"We carried out a nondeterministic analysis of flow in a fluidized bed. The flow in the fluidized bed is simulated with National Energy Technology Laboratory’s open-source multiphase fluid dynamics suite MFiX. It does not possess tools for uncertainty quantification. Therefore, we developed a C++ wrapper to integrate an uncertainty quantification toolkit developed at Sandia National Laboratory with MFiX. The wrapper exchanges uncertain input parameters and key output parameters among Dakota and MFiX. However, a data-driven framework is also developed to obtain reliable statistics as it is not feasible to get them with MFiX integrated into Dakota, Dakota-MFiX. The data generated from Dakota-MFiX simulations, with the Latin Hypercube method of sampling size 500, is used to train a machine-learning algorithm. The trained and tested deep neural network algorithm is integrated with Dakota via the wrapper to obtain low order statistics of the bed height and pressure drop across the bed.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126909919","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}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916378
J. Ellis, S. Rajamanickam
Over the last decade, hardware advances have led to the feasibility of training and inference for very large deep neural networks. Sparsified deep neural networks (DNNs) can greatly reduce memory costs and increase throughput of standard DNNs, if loss of accuracy can be controlled. The IEEE HPEC Sparse Deep Neural Network Graph Challenge serves as a testbed for algorithmic and implementation advances to maximize computational performance of sparse deep neural networks. We base our sparse network for DNNs, KK-SpDNN, on the sparse linear algebra kernels within the Kokkos Kernels library. Using the sparse matrix-matrix multiplication in Kokkos Kernels allows us to reuse a highly optimized kernel. We focus on reducing the single node and multi-node runtimes for 12 sparse networks. We test KK-SpDNN on Intel Skylake and Knights Landing architectures and see 120-500x improvement on single node performance over the serial reference implementation. We run in data-parallel mode with MPI to further speed up network inference, ultimately obtaining an edge processing rate of 1.16e+12 on 20 Skylake nodes. This translates to a 13x speed up on 20 nodes compared to our highly optimized multithreaded implementation on a single Skylake node.
{"title":"Scalable Inference for Sparse Deep Neural Networks using Kokkos Kernels","authors":"J. Ellis, S. Rajamanickam","doi":"10.1109/HPEC.2019.8916378","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916378","url":null,"abstract":"Over the last decade, hardware advances have led to the feasibility of training and inference for very large deep neural networks. Sparsified deep neural networks (DNNs) can greatly reduce memory costs and increase throughput of standard DNNs, if loss of accuracy can be controlled. The IEEE HPEC Sparse Deep Neural Network Graph Challenge serves as a testbed for algorithmic and implementation advances to maximize computational performance of sparse deep neural networks. We base our sparse network for DNNs, KK-SpDNN, on the sparse linear algebra kernels within the Kokkos Kernels library. Using the sparse matrix-matrix multiplication in Kokkos Kernels allows us to reuse a highly optimized kernel. We focus on reducing the single node and multi-node runtimes for 12 sparse networks. We test KK-SpDNN on Intel Skylake and Knights Landing architectures and see 120-500x improvement on single node performance over the serial reference implementation. We run in data-parallel mode with MPI to further speed up network inference, ultimately obtaining an edge processing rate of 1.16e+12 on 20 Skylake nodes. This translates to a 13x speed up on 20 nodes compared to our highly optimized multithreaded implementation on a single Skylake node.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"168 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115584051","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}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916236
Arnab A. Purkayastha, S. Raghavendran, Jhanani Thiagarajan, H. Tabkhi
OpenCL programming ability combined with OpenCL High-Level Synthesis (OpenCL-HLS) tools have made tremendous improvements in the reconfigurable computing field. FPGAs inherent pipelined parallelism capability provides not only faster execution times but also power-efficient solutions when executing massively parallel applications. A major execution bottleneck affecting FPGA performance is the high number of memory stalls exposed to pipelined data-path that hinders the benefits of data-path customization.This paper explores the efficiency of “OpenCL Pipe” to hide memory access latency on cloud FPGAs by decoupling memory access from computation. The Pipe semantic is leveraged to split OpenCL kernels into “read”, “compute” and “write back” sub-kernels which work concurrently to overlap the computation of current threads with the memory access of future threads. For evaluation, we use a mix of seven massively parallel high-performance applications from the Rodinia suite vs. 3.1. All our tests are conducted on the Xilinx VU9FP FPGA platform of Amazon cloud-based AWS EC2 F1 instance. On average, we observe 5.2x speedup with a 2.2x increase in memory bandwidth utilization with about 2.5x increase in FPGA resource utilization over the baseline synthesis (Xilinx OpenCL-HLS).11This work has been funded and supported by the Xilinx University Program (XUP)..
{"title":"Exploring the Efficiency of OpenCL Pipe for Hiding Memory Latency on Cloud FPGAs","authors":"Arnab A. Purkayastha, S. Raghavendran, Jhanani Thiagarajan, H. Tabkhi","doi":"10.1109/HPEC.2019.8916236","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916236","url":null,"abstract":"OpenCL programming ability combined with OpenCL High-Level Synthesis (OpenCL-HLS) tools have made tremendous improvements in the reconfigurable computing field. FPGAs inherent pipelined parallelism capability provides not only faster execution times but also power-efficient solutions when executing massively parallel applications. A major execution bottleneck affecting FPGA performance is the high number of memory stalls exposed to pipelined data-path that hinders the benefits of data-path customization.This paper explores the efficiency of “OpenCL Pipe” to hide memory access latency on cloud FPGAs by decoupling memory access from computation. The Pipe semantic is leveraged to split OpenCL kernels into “read”, “compute” and “write back” sub-kernels which work concurrently to overlap the computation of current threads with the memory access of future threads. For evaluation, we use a mix of seven massively parallel high-performance applications from the Rodinia suite vs. 3.1. All our tests are conducted on the Xilinx VU9FP FPGA platform of Amazon cloud-based AWS EC2 F1 instance. On average, we observe 5.2x speedup with a 2.2x increase in memory bandwidth utilization with about 2.5x increase in FPGA resource utilization over the baseline synthesis (Xilinx OpenCL-HLS).11This work has been funded and supported by the Xilinx University Program (XUP)..","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131352823","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}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916302
Seher Acer, Abdurrahman Yasar, S. Rajamanickam, Michael M. Wolf, Ümit V. Çatalyürek
Triangle counting is a foundational graph-analysis kernel in network science. It has also been one of the challenge problems for the “Static Graph Challenge”. In this work, we propose a novel, hybrid, parallel triangle counting algorithm based on its linear algebra formulation. Our framework uses MPI and Cilk to exploit the benefits of distributed-memory and shared-memory parallelism, respectively. The problem is partitioned among MPI processes using a two-dimensional (2D) Cartesian block partitioning. One-dimensional (1D) rowwise partitioning is used within the Cartesian blocks for shared-memory parallelism using the Cilk programming model. Besides exhibiting very good strong scaling behavior in almost all tested graphs, our algorithm achieves the fastest time on the 1.4B edge real-world twitter graph, which is 3.217 seconds, on 1,092 cores. In comparison to past distributed-memory parallel winners of the graph challenge, we demonstrate a speed up of 2.7× on this twitter graph. This is also the fastest time reported for parallel triangle counting on the twitter graph when the graph is not replicated.
{"title":"Scalable Triangle Counting on Distributed-Memory Systems","authors":"Seher Acer, Abdurrahman Yasar, S. Rajamanickam, Michael M. Wolf, Ümit V. Çatalyürek","doi":"10.1109/HPEC.2019.8916302","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916302","url":null,"abstract":"Triangle counting is a foundational graph-analysis kernel in network science. It has also been one of the challenge problems for the “Static Graph Challenge”. In this work, we propose a novel, hybrid, parallel triangle counting algorithm based on its linear algebra formulation. Our framework uses MPI and Cilk to exploit the benefits of distributed-memory and shared-memory parallelism, respectively. The problem is partitioned among MPI processes using a two-dimensional (2D) Cartesian block partitioning. One-dimensional (1D) rowwise partitioning is used within the Cartesian blocks for shared-memory parallelism using the Cilk programming model. Besides exhibiting very good strong scaling behavior in almost all tested graphs, our algorithm achieves the fastest time on the 1.4B edge real-world twitter graph, which is 3.217 seconds, on 1,092 cores. In comparison to past distributed-memory parallel winners of the graph challenge, we demonstrate a speed up of 2.7× on this twitter graph. This is also the fastest time reported for parallel triangle counting on the twitter graph when the graph is not replicated.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133160698","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}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916216
Chuangyi Gui, Long Zheng, Pengcheng Yao, Xiaofei Liao, Hai Jin
Triangle counting is one of the most basic graph applications to solve many real-world problems in a wide variety of domains. Exploring the massive parallelism of the Graphics Processing Unit (GPU) to accelerate the triangle counting is prevail. We identify that the stat-of-the-art GPU-based studies that focus on improving the load balancing still exhibit inherently a large number of random accesses in degrading the performance. In this paper, we design a prefetching scheme that buffers the neighbor list of the processed vertex in advance in the fast shared memory to avoid high latency of random global memory access. Also, we adopt the degree-based graph reordering technique and design a simple heuristic to evenly distribute the workload. Compared to the state-of-the-art HEPC Graph Challenge Champion in the last year, we advance to improve the performance of triangle counting by up to $5.9 times $ speedup with $gt 10^{9}$ TEPS on a single GPU for many large real graphs from graph challenge datasets.
{"title":"Fast Triangle Counting on GPU","authors":"Chuangyi Gui, Long Zheng, Pengcheng Yao, Xiaofei Liao, Hai Jin","doi":"10.1109/HPEC.2019.8916216","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916216","url":null,"abstract":"Triangle counting is one of the most basic graph applications to solve many real-world problems in a wide variety of domains. Exploring the massive parallelism of the Graphics Processing Unit (GPU) to accelerate the triangle counting is prevail. We identify that the stat-of-the-art GPU-based studies that focus on improving the load balancing still exhibit inherently a large number of random accesses in degrading the performance. In this paper, we design a prefetching scheme that buffers the neighbor list of the processed vertex in advance in the fast shared memory to avoid high latency of random global memory access. Also, we adopt the degree-based graph reordering technique and design a simple heuristic to evenly distribute the workload. Compared to the state-of-the-art HEPC Graph Challenge Champion in the last year, we advance to improve the performance of triangle counting by up to $5.9 times $ speedup with $gt 10^{9}$ TEPS on a single GPU for many large real graphs from graph challenge datasets.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"114 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133540336","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}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916298
M. Rogowski, Suha N. Kayum, F. Mannuß
In parallel reservoir simulation, massively sized files are written recurrently throughout a simulation run. A method is developed to compress the distributed data to be written during the simulation run and to output it to a single compressed file. Evaluation of several compression algorithms on a range of simulation models is performed. The presented method results in 3x file size reduction and a decrease in the total application runtime.
{"title":"Lossless Compression of Internal Files in Parallel Reservoir Simulation","authors":"M. Rogowski, Suha N. Kayum, F. Mannuß","doi":"10.1109/HPEC.2019.8916298","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916298","url":null,"abstract":"In parallel reservoir simulation, massively sized files are written recurrently throughout a simulation run. A method is developed to compress the distributed data to be written during the simulation run and to output it to a single compressed file. Evaluation of several compression algorithms on a range of simulation models is performed. The presented method results in 3x file size reduction and a decrease in the total application runtime.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"79 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132998017","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}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916528
A. Wollaber, Jaime Peña, Benjamin Blease, Leslie Shing, Kenneth Alperin, Serge Vilvovsky, P. Trepagnier, Neal Wagner, Leslie Leonard
Cyber situation awareness technologies have largely been focused on present-state conditions, with limited abilities to forward-project nominal conditions in a contested environment. We demonstrate an approach that uses data-driven, high performance computing (HPC) simulations of attacker/defender activities in a logically connected network environment that enables this capability for interactive, operational decision making in real time. Our contributions are three-fold: (1) we link live cyber data to inform the parameters of a cybersecurity model, (2) we perform HPC simulations and optimizations with a genetic algorithm to evaluate and recommend risk remediation strategies that inhibit attacker lateral movement, and (3) we provide a prototype platform to allow cyber defenders to assess the value of their own alternative risk reduction strategies on a relevant timeline. We present an overview of the data and software architectures, and results are presented that demonstrate operational utility alongside HPC-enabled runtimes.
{"title":"Proactive Cyber Situation Awareness via High Performance Computing","authors":"A. Wollaber, Jaime Peña, Benjamin Blease, Leslie Shing, Kenneth Alperin, Serge Vilvovsky, P. Trepagnier, Neal Wagner, Leslie Leonard","doi":"10.1109/HPEC.2019.8916528","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916528","url":null,"abstract":"Cyber situation awareness technologies have largely been focused on present-state conditions, with limited abilities to forward-project nominal conditions in a contested environment. We demonstrate an approach that uses data-driven, high performance computing (HPC) simulations of attacker/defender activities in a logically connected network environment that enables this capability for interactive, operational decision making in real time. Our contributions are three-fold: (1) we link live cyber data to inform the parameters of a cybersecurity model, (2) we perform HPC simulations and optimizations with a genetic algorithm to evaluate and recommend risk remediation strategies that inhibit attacker lateral movement, and (3) we provide a prototype platform to allow cyber defenders to assess the value of their own alternative risk reduction strategies on a relevant timeline. We present an overview of the data and software architectures, and results are presented that demonstrate operational utility alongside HPC-enabled runtimes.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"32 Pt 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133993922","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}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916224
Cole Campton, Xiaobai Sun
We establish first a theoretical foundation for the use of Gromov-Hausdorff (GH) distance for point set registration with homeomorphic deformation maps perturbed by Gaussian noise. We then present a probabilistic, deformable registration framework. At the core of the framework is a highly efficient iterative algorithm with guaranteed convergence to a local minimum of the GH-based objective function. The framework has two other key components – a multi-scale stochastic shape descriptor and a data compression scheme. We also present an experimental comparison between our method and two existing influential methods on non-rigid motion between digital anthropomorphic phantoms extracted from physical data of multiple individuals.
{"title":"Auxiliary Maximum Likelihood Estimation for Noisy Point Cloud Registration","authors":"Cole Campton, Xiaobai Sun","doi":"10.1109/HPEC.2019.8916224","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916224","url":null,"abstract":"We establish first a theoretical foundation for the use of Gromov-Hausdorff (GH) distance for point set registration with homeomorphic deformation maps perturbed by Gaussian noise. We then present a probabilistic, deformable registration framework. At the core of the framework is a highly efficient iterative algorithm with guaranteed convergence to a local minimum of the GH-based objective function. The framework has two other key components – a multi-scale stochastic shape descriptor and a data compression scheme. We also present an experimental comparison between our method and two existing influential methods on non-rigid motion between digital anthropomorphic phantoms extracted from physical data of multiple individuals.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134328438","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}
Pub Date : 2019-09-01DOI: 10.1109/HPEC.2019.8916572
Chunxing Yin, E. J. Riedy
Most current frameworks for streaming graph analysis “stop the world” and halt ingesting data while updating analysis results. Others add overhead for different forms of version control. In both methods, adding additional analysis kernels adds additional overhead to the entire system. A new formal model of concurrent analysis lets some algorithms, those valid for the model, update results concurrently with data ingest without synchronization. Additional kernels incur very little overhead. Here we present the first experimental results for the new model, considering the performance and result latency of updating Katz centrality on a low-power edge platform. The Katz centrality values remain close to the synchronous algorithm while reducing latency delay from 12.8$times $ to 179$times $.
{"title":"Concurrent Katz Centrality for Streaming Graphs","authors":"Chunxing Yin, E. J. Riedy","doi":"10.1109/HPEC.2019.8916572","DOIUrl":"https://doi.org/10.1109/HPEC.2019.8916572","url":null,"abstract":"Most current frameworks for streaming graph analysis “stop the world” and halt ingesting data while updating analysis results. Others add overhead for different forms of version control. In both methods, adding additional analysis kernels adds additional overhead to the entire system. A new formal model of concurrent analysis lets some algorithms, those valid for the model, update results concurrently with data ingest without synchronization. Additional kernels incur very little overhead. Here we present the first experimental results for the new model, considering the performance and result latency of updating Katz centrality on a low-power edge platform. The Katz centrality values remain close to the synchronous algorithm while reducing latency delay from 12.8$times $ to 179$times $.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"120 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114210157","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：