9,394 Hits in 4.7 sec

Big Graph Analytics Platforms

Da Yan, Yingyi Bu, Yuanyuan Tian, Amol Deshpande
2017 Foundations and Trends in Databases  
algebra operations.  ...  Furthermore, SystemML supports general linear algebra operations, besides just matrix-vector multiplication, and is thus applicable for any ML or analytics algorithms that can be expressed using linear  ...  Then, parallel asynchronous I/O requests are submitted to the SSD for those required pages that are not in the buffer pool.  ... 
doi:10.1561/1900000056 fatcat:ucqrtzo4q5g2lpj6dmp7jv3e5m

RHEEM: enabling cross-platform data processing

Divy Agrawal, Mourad Ouzzani, Paolo Papotti, Jorge-Arnulfo Quiané-Ruiz, Nan Tang, Saravanan Thirumuruganathan, Anis Troudi, Sanjay Chawla, Bertty Contreras-Rojas, Ahmed Elmagarmid, Yasser Idris, Zoi Kaoudi (+3 others)
2018 Proceedings of the VLDB Endowment  
It features (i) a robust interface to easily compose data analytic tasks; (ii) a novel cost-based optimizer able to find the most efficient platform in almost all cases; and (iii) an executor to efficiently  ...  Using different real-world applications with Rheem, we demonstrate how cross-platform data processing can accelerate performance by more than one order of magnitude compared to single-platform data processing  ...  In the former case, we aim at seamlessly integrating all data analytic activity governing an aircraft; In the latter case, we aim at reducing the effort scientists need for building data analytic pipelines  ... 
doi:10.14778/3236187.3236195 fatcat:xt4qxbumhjfptfb3g2xkkkjlay

A mixed-platform framework for Dynamic Stability Assessment

T. Kyriakidis, G. Lanz, D. Sallin, G. Lilis, L. Fabre, R. Cherkaoui, M. Kayal
2013 2013 IEEE Power & Energy Society General Meeting  
Data flow control, user interfacing, configuration, result post-processing and other auxiliary operations are realized in software.  ...  In this work, operations are shared between a software platform and a hardware one. TD simulation is handled by a dedicated mixed-signal electronics implementation.  ...  The protocol used to communicate commands and data between the two platforms is a flag-polled shared memory scheme.  ... 
doi:10.1109/pesmg.2013.6672431 fatcat:4gi36kah5fcz5eqjoghe2izpau

Development of A Scalable Platform for Large-scale Reservoir Simulations on Parallel computers [article]

Hui Liu, Kun Wang, Bo Yang, Zhangxin Chen
2018 arXiv   pre-print
This paper presents our work on designing a parallel platform for large-scale reservoir simulations.  ...  It also has an interface to a parallel algebraic multigrid solver, BoomerAMG from HYPRE.  ...  Algebraic Operations With the help of above data structures, commonly used matrix-vector operations and vector operations can be implemented directly, which are listed as follows: y = αAx + βy, (14) z  ... 
arXiv:1602.05901v6 fatcat:js77vehwjnbj7b644f7byxebpu

Overview of Platform Architectures [chapter]

Alexander Supalov, Andrey Semin, Michael Klemm, Christopher Dahnken
2014 Optimizing HPC Applications with Intel® Cluster Tools  
at 2.5 GHz), up to around 250 ns for the loaded case (over 600 cycles).  ...  The best efficiencies close to 95 percent are usually obtained by highly tuned computational kernels, such as BLAS (Basic Linear Algebra Subprograms), while mainstream HPC applications often achieve efficiencies  ... 
doi:10.1007/978-1-4302-6497-2_2 fatcat:fl6uujc7lbcu3dyayluynxpmlu

GPU-Accelerated Evaluation Platform for High Fidelity Network Modeling

Zhiguo Xu, Rajive Bagrodia
2007 21st International Workshop on Principles of Advanced and Distributed Simulation (PADS'07)  
This benefit is particularly impressive for models that include significant data parallel computations.  ...  This study suggests that besides parallel computing and grid computing, network simulations can also be scaled by reaping computation capability of GPUs and, potentially, other external computational hardware  ...  First, analytical approximation of network (traffic) dynamics may be modeled with a set of linear algebra operations, which can effectively exploit the data parallelism of a GPU.  ... 
doi:10.1109/pads.2007.20 dblp:conf/pads/XuB07 fatcat:hsp52mtrh5cptm5p2lmokcbhs4

Java simulation platform for control system based on block diagram

Masanobu Koga, Yusuke Tsutsui, Jun Yabuuchi
2006 2006 IEEE Conference on Computer Aided Control System Design, 2006 IEEE International Conference on Control Applications, 2006 IEEE International Symposium on Intelligent Control  
A Java simulation platform, which has been developed based on the proposed method, is introduced. It provides an interactive graphical environment for modeling and simulation of control systems.  ...  Block Diagram of Control System structure of the adjacency matrix is useful for the data operation in computer. 1 S 2 S 3 S  ...  They give the following two algorithms to get the linearized approximate model. 1) numerical-perturbation linearization 2) block-by-block analytic linearization The former linearizes the whole system by  ... 
doi:10.1109/cacsd-cca-isic.2006.4776999 fatcat:gamivytdkfacxn6zggg3fygtyi

The Stratosphere platform for big data analytics

Alexander Alexandrov, Rico Bergmann, Stephan Ewen, Johann-Christoph Freytag, Fabian Hueske, Arvid Heise, Odej Kao, Marcus Leich, Ulf Leser, Volker Markl, Felix Naumann, Mathias Peters (+6 others)
2014 The VLDB journal  
We present Stratosphere, an open-source software stack for parallel data analysis.  ...  Stratosphere brings together a unique set of features that allow the expressive, easy, and efficient programming of analytical applications at very large scale.  ...  Query languages and models for parallel data management It is conceptually easy to parallelize the basic operators of relational algebra, and parallel analytical databases have existed for decades.  ... 
doi:10.1007/s00778-014-0357-y fatcat:ficnpssbvjdatjs3gewa4jdd7m

Benchmarking Parallelism in FaaS Platforms [article]

Daniel Barcelona-Pons, Pedro García-López
2021 arXiv   pre-print
Some systems tackle Function-as-a-Service (FaaS) properties on automatic elasticity and scale to run highly-parallel computing jobs.  ...  However, they focus on specific platforms and convey that their ideas can be extrapolated to any FaaS runtime. An important question arises: do all FaaS platforms fit parallel computations?  ...  This work has been partially supported by the EU project H2020 "CloudButton: Serverless Data Analytics Platform" (825184) and by the Spanish government (PID2019-106774RB-C22).  ... 
arXiv:2010.15032v3 fatcat:cogt6e7uinehveca3fmh4kuilm

The ViP2P Platform: XML Views in P2P [article]

Konstantinos Karanasos, Asterios Katsifodimos, Ioana Manolescu (INRIA Saclay - Ile de France, LRI), Spyros Zoupanos (INRIA Saclay - Ile de France, LRI, MPII)
2011 arXiv   pre-print
The growing volumes of XML data sources on the Web or produced by enterprises, organizations etc. raise many performance challenges for data management applications.  ...  ViP2P innovates over other similar DHT-based XML sharing platforms by using a very expressive structured XML query language.  ...  We observe that parallel extraction accelerates data extraction (in this case, up to 40%).  ... 
arXiv:1112.2610v1 fatcat:vymszw2infhxddznapyur65hya

Enabling particle applications for exascale computing platforms

Susan M Mniszewski, James Belak, Jean-Luc Fattebert, Christian FA Negre, Stuart R Slattery, Adetokunbo A Adedoyin, Robert F Bird, Choongseok Chang, Guangye Chen, Stéphane Ethier, Shane Fogerty, Salman Habib (+12 others)
2021 The international journal of high performance computing applications  
., electrostatic MD and gravitational N-body), particle-in-cell (PIC) methods, and linear-scaling electronic structure and quantum molecular dynamics (QMD) algorithms.  ...  Proxy apps are vehicles used to evaluate the viability of incorporating various types of algorithms, data structures, and architecture-specific optimizations and the associated trade-offs; examples include  ...  Loops over nonzero matrix elements are replaced by loops over nonzero blocks, and dense linear algebra operations are done on blocks.  ... 
doi:10.1177/10943420211022829 fatcat:7goiighkmjgjnosvzjubskjplq

Benchmarking parallelism in FaaS platforms

Daniel Barcelona-Pons, Pedro García-López
2021 Future generations computer systems  
Some systems tackle Functionas-a-Service (FaaS) properties on automatic elasticity and scale to run highly-parallel computing jobs.  ...  However, they focus on specific platforms and convey that their ideas can be extrapolated to any FaaS runtime. An important question arises: do all FaaS platforms fit parallel computations?  ...  This work has been partially supported by the EU project H2020 ''CloudButton: Serverless Data Analytics Platform'' (825184) and by the Spanish government (PID2019-106774RB-C22).  ... 
doi:10.1016/j.future.2021.06.005 fatcat:fogjo5e3sjborg3bbfkupoihq4

Prediction of Time-to-Solution in Material Science Simulations Using Deep Learning

Federico Pittino, Pietro Bonfà, Andrea Bartolini, Fabio Affinito, Luca Benini, Carlo Cavazzoni
2019 Proceedings of the Platform for Advanced Scientific Computing Conference on - PASC '19  
Predicting the time to solution for massively parallel scientific codes is a complex task.  ...  A reliable prediction of execution time is however of great importance to the user who wants to plan on large scale simulations or virtual screening procedures characteristic of high throughput computing  ...  For the basic linear algebra operations we parametrize the performance of the implementations of the DGEMM and the ZGEMM subroutines.  ... 
doi:10.1145/3324989.3325720 dblp:conf/pasc/PittinoBBABC19 fatcat:n4x2owebmrgs7coudflrzkji4e

An optimizing multi-platform source-to-source compiler framework for the NEURON MODeling Language [article]

Pramod Kumbhar, Omar Awile, Liam Keegan, Jorge Blanco Alonso, James King, Michael Hines, Felix Schürmann
2019 arXiv   pre-print
Benchmarks were performed on Intel Skylake, Intel KNL and AMD Naples platforms.  ...  Here, we describe a new code generation framework (NMODL) for an existing DSL in the NEURON framework, a widely used software for massively parallel simulation of biophysically detailed brain tissue models  ...  While excellent scaling has been achieved on a variety of platforms with the conversion to Single Instruction Multiple Data (SIMD) implementations, domain specific knowledge expressed in the models is  ... 
arXiv:1905.02241v1 fatcat:3wcl7sasiffehc65rp3cabbuym

Heterogeneous parallel computing: from clusters of workstations to hierarchical hybrid platforms

2014 Supercomputing Frontiers and Innovations  
The paper overviews the state of the art in design and implementation of data parallel scientific applications on heterogeneous platforms.  ...  both traditional approaches originally designed for clusters of heterogeneous workstations and the most recent methods developed in the context of modern multicore and multi-accelerator heterogeneous platforms  ...  Matrix partitioning problems occur during the design of parallel linear algebra algorithms for heterogeneous platforms.  ... 
doi:10.14529/jsfi140304 fatcat:f4u6c4zupjhvlg5c666shbxto4
« Previous Showing results 1 — 15 out of 9,394 results