535 Hits in 7.8 sec

Dynamic Model-Driven Parallel I/O Performance Tuning

Babak Behzad, Surendra Byna, Stefan M. Wild, Prabhat, Marc Snir
2015 2015 IEEE International Conference on Cluster Computing  
Using a set of real measurements from running an I/O kernel as training set, we generate a nonlinear regression model.  ...  For instance, analytical models fail to capture the dynamic nature of shared supercomputing systems and are application-specific.  ...  Development of Empirical Performance Models We now examine nonlinear regression models in the context of modeling I/O write times for a given application.  ... 
doi:10.1109/cluster.2015.37 dblp:conf/cluster/BehzadBWPS15 fatcat:2e5sg4jbxvhu5ceclfxdinddxm

Improving parallel I/O autotuning with performance modeling

Babak Behzad, Surendra Byna, Stefan M. Wild, Mr. Prabhat, Marc Snir
2014 Proceedings of the 23rd international symposium on High-performance parallel and distributed computing - HPDC '14  
We are working towards an autotuning framework for determining the parallel I/O parameters that can achieve good I/O performance for different data write patterns.  ...  In this paper, we characterize parallel I/O and discuss the development of predictive models for use in effectively reducing the parameter space.  ...  We characterize performance of a typical parallel I/O subsystem with multiple levels of data movement and develop performance prediction models.  ... 
doi:10.1145/2600212.2600708 dblp:conf/hpdc/BehzadBWPS14 fatcat:k5u34mmuabd7zgjcjaqfd64q74

Predicting Performance of Non-contiguous I/O with Machine Learning [chapter]

Julian Kunkel, Michaela Zimmer, Eugen Betke
2015 Lecture Notes in Computer Science  
Data sieving in ROMIO promises to optimize individual noncontiguous I/O.  ...  Additionally, we demonstrate the versatility of this approach by applying it to the porting system of DKRZ's next generation supercomputer and discuss achievable performance gains.  ...  Related Work Widely used concepts to improve I/O performance are non-blocking I/O, data pre-fetching and write-behind.  ... 
doi:10.1007/978-3-319-20119-1_19 fatcat:mxbdqfmicvg5pmbdvwzxgh52uy

Toward Understanding I/O Behavior in HPC Workflows

Jakob Luttgau, Shane Snyder, Philip Carns, Justin M. Wozniak, Julian Kunkel, Thomas Ludwig
2018 2018 IEEE/ACM 3rd International Workshop on Parallel Data Storage & Data Intensive Scalable Computing Systems (PDSW-DISCS)  
In this paper, we present an approach to augment the I/O efficiency of the individual tasks of workflows by combining workflow description frameworks with system I/O telemetry data.  ...  I/O performance is not usually accounted for in workflow telemetry reported to users.  ...  I/O Perspective on Scientific Workflows Scientists commonly perform experiments on supercomputers using some notion of a workflow.  ... 
doi:10.1109/pdsw-discs.2018.00012 dblp:conf/sc/LuttgauSCWK018 fatcat:rjlmvxiefbfrznlcu4p226kdwu

Understanding I/O workload characteristics of a Peta-scale storage system

Youngjae Kim, Raghul Gunasekaran
2014 Journal of Supercomputing  
We also study the I/O load imbalance problems using I/O performance data collected from the Spider storage system.  ...  From this study, we develop synthesized workloads, and we show that the read and write I/O bandwidth usage as well as the inter-arrival time of requests can be modeled as a Pareto distribution.  ...  Figure 7 presents the percentage of write requests with respect to the total number of I/O requests in the system.  ... 
doi:10.1007/s11227-014-1321-8 fatcat:ww4xluintffsniif4t2hrxupea

Informed Prefetching and Caching [chapter]

2009 High Performance Mass Storage and Parallel I/O  
The underutilization of disk parallelism and file cache buffers by traditional file systems induces I/O stall time that degrades the performance of modern microprocessor-based systems.  ...  In this paper, we present aggressive mechanisms that tailor file system resource management to the needs of I/O-intensive applications.  ...  for help with the Davidson algorithm, Ravi Mosur and the Sphinx group, here at CMU, Jiawen Su, who did the initial port of TIP to OSF/1 from Mach, David Golub for his debugging and coding contributions  ... 
doi:10.1109/9780470544839.ch16 fatcat:xn47z46ga5asfanpxyxwcfyyy4

ViPIOS - VIenna Parallel Input Output System: Language, Compiler and Advanced Data Structure Support for Parallel I/O Operations [article]

Erich Schikuta, Helmut Wanek, Heinz Stockinger, Kurt Stockinger, Thomas Fürle, Oliver Jorns, Christoph Löffelhardt, Peter Brezany, Minh Dang, Thomas Mück
2018 arXiv   pre-print
We focus on the design of an advanced parallel I/O support, called ViPIOS (VIenna Parallel I/O System), to be targeted by language compilers supporting the same programming model like High Performance  ...  Based on this formal model we designed and developed an I/O runtime system, ViPIOS, which provides support for several issues, as - parallel access to files for read/write operations, - optimization of  ...  The appendix gives a detailed survey of all the groups and systems mentioned in the Dictionary part and compares the features.  ... 
arXiv:1808.01166v1 fatcat:gcfewrlk5fcqlbdztveocldqim

Performance degradation analysis of a supercomputer

Domenico Cotroneo, Flavio Frattini, Roberto Natella, Roberto Pietrantuono
2013 2013 IEEE International Symposium on Software Reliability Engineering Workshops (ISSREW)  
We analyze performance degradation phenomena due to software aging on a real supercomputer deployed at the Federico II University of Naples, by considering a dataset of ten months of operational usage.  ...  The analysis pinpointed performance degradation trends that were actually caused by the gradual error accumulation within basic software of the supercomputer.  ...  I/O-bound jobs.  ... 
doi:10.1109/issrew.2013.6688904 dblp:conf/issre/CotroneoFNP13 fatcat:3ekubhqvjrgstjqvknzdf6lncq

Understanding Metadata Latency with MDWorkbench [chapter]

Julian Martin Kunkel, George S. Markomanolis
2018 Lecture Notes in Computer Science  
While parallel file systems often satisfy the need of applications with bulk synchronous I/O, they lack capabilities of dealing with metadata intense workloads.  ...  However, metadata performance is crucial for interactive use. Metadata benchmarks involve even more parameters compared to I/O benchmarks.  ...  In the second phase, the I/O configuration analysis, factors that impact the I/O system performance are identified, such as file system and I/O node connection.  ... 
doi:10.1007/978-3-030-02465-9_5 fatcat:hjkkftr2mbd7fakkgayigp4ox4

ICT Asset Prices: Marshaling Evidence into New Measures

David Byrne, Carol Corrado
2017 Finance and Economics Discussion Series  
It provides the sources and methods used to construct national accounts-style price deflators for the major components of ICT investment-communications equipment, computer equipment, and software-that  ...  The table is structured according to I-O item codes, which largely align with NAICS industry codes.  ...  As in the case of communications equipment, the table is structured according to I-O item codes, which largely align with NAICS industry codes.  ... 
doi:10.17016/feds.2017.016r1 fatcat:erdtkyg5ubh6dpfkndkosv34ei


Mingliang Liu, Ye Jin, Jidong Zhai, Yan Zhai, Qianqian Shi, Xiaosong Ma, Wenguang Chen
2013 Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis on - SC '13  
It has been observed that changing cloud I/O system configurations leads to significant variation in the performance and cost efficiency of I/O intensive HPC applications.  ...  This new environment highlights the I/O bottleneck problem, typically with top-of-the-line compute instances but sub-par communication and I/O facilities.  ...  Prediction model Many studies exist on performance modeling for HPC applications and/or I/O systems [19, 46, 52, 54] .  ... 
doi:10.1145/2503210.2503216 dblp:conf/sc/LiuJZZSMC13 fatcat:6p7d3z5ttbbp5ilpjdw5imsd3a

Evaluating the performance of parallel subsurface simulators: An illustrative example with PFLOTRAN

G. E. Hammond, P. C. Lichtner, R. T. Mills
2014 Water Resources Research  
code PFLOTRAN as it is applied to several realistic modeling scenarios run on the Jaguar supercomputer.  ...  Weak scalability is discussed in detail for the regional doublet problem, and several difficulties with its interpretation are noted.  ...  parallel I/O with the two-stage I/O option provided by collaborators [Sripathi, 2010] .  ... 
doi:10.1002/2012wr013483 pmid:25506097 pmcid:PMC4257577 fatcat:cdwcpfmkhjgv5nofkavjmwcasm

Performance Evaluation of the PVFS2 Architecture

Julian M. Kunkel, Thomas Ludwig
2007 15th EUROMICRO International Conference on Parallel, Distributed and Network-Based Processing (PDP'07)  
Also, the influence of the hardware components on the performance is modeled to estimate the maximum achievable performance of a parallel file system.  ...  The paper focusses on the Parallel Virtual File System 2 (PVFS2) and shows results for the functionality file creation, small contiguous I/O requests and large contiguous I/O requests.  ...  In order to improve the implementation of a parallel file system developers determine the performance with I/O benchmarks.  ... 
doi:10.1109/pdp.2007.65 dblp:conf/pdp/Kunkel007 fatcat:wgkstefl7jdv7g6mvjzzzh7xzi

Can parallelization save the (computing) world?

János Végh, József Vásárhelyi, Dániel Drótos
2019 Advances in Science, Technology and Engineering Systems  
The model enables to highlight bottlenecks of parallelized systems of different kind and with the published performance data enables to predict performance limits of strongly parallelized systems like  ...  Some general limitations are shortly discussed, and a (by intention strongly simplified) general model of performance of parallelized systems is introduced.  ...  Acknowledgment Project no. 125547 has been implemented with the support provided from the National Research, Development and Innovation Fund of Hungary, financed under the K funding scheme.  ... 
doi:10.25046/aj040114 fatcat:i73yfkiy4jefbewg3y4m7jifqi

Artificial Intelligence: An Energy Efficiency Tool for Enhanced High performance computing

Anabi Hilary Kelechi, Mohammed H. Alsharif, Okpe Jonah Bameyi, Paul Joan Ezra, Iorshase Kator Joseph, Aaron-Anthony Atayero, Zong Woo Geem, Junhee Hong
2020 Symmetry  
Today's high level of computing power from supercomputers comes at the expense of consuming large amounts of electric power.  ...  Power-consuming entities such as high performance computing (HPC) sites and large data centers are growing with the advance in information technology.  ...  DVFS approaches have been leveraged to reduce depredating performance by addressing various application characteristics, including communication barriers, I/O delays, load imbalances, or repetitive behaviors  ... 
doi:10.3390/sym12061029 fatcat:3ll5wwfrhnhwlgoua2qpi5kb34
« Previous Showing results 1 — 15 out of 535 results