Filters








6,027 Hits in 8.1 sec

Green Power Constrained Scheduling for Sequential Independent Tasks on Identical Parallel Machines

Laurent Philippe, Jean-Marc Nicod, Laurent Philippe, Veronika Rehn-Sonigo
2019 2019 IEEE Intl Conf on Parallel & Distributed Processing with Applications, Big Data & Cloud Computing, Sustainable Computing & Communications, Social Computing & Networking (ISPA/BDCloud/SocialCom/SustainCom)  
We tackle here the problem of scheduling independent tasks on a multi-machine platform that is exclusively run with green energy.  ...  We propose different power constrained scheduling algorithms, and evaluate them through an experimental study on an HPC model that considers the possibility to switch machines on or off.  ...  Computations have been performed on the supercomputer facilities of the Mésocentre de calcul de Franche-Comté -Besançon.  ... 
doi:10.1109/ispa-bdcloud-sustaincom-socialcom48970.2019.00029 dblp:conf/ispa/KassabN0R19 fatcat:v2jijt6x7jdbpnq2cqtnfw6zb4

Assessing the Use of Genetic Algorithms to Schedule Independent Tasks Under Power Constraints

Ayham Kassab, Jean-Marc Nicod, Laurent Philippe, Veronika Rehn-Sonigo
2018 2018 International Conference on High Performance Computing & Simulation (HPCS)  
In this paper, we present genetic algorithms for scheduling sets of independent tasks in parallel, with the objective of minimizing the makespan under power availability constraints.  ...  The optimization problem of scheduling a set of tasks under power constraints is however proven to be NP-Complete.  ...  Computations have been performed on the supercomputer facilities of the Mésocentre de calcul de Franche-Comté -Besançon.  ... 
doi:10.1109/hpcs.2018.00052 dblp:conf/ieeehpcs/KassabN0R18 fatcat:b6bgszfj7rhpzgg37nscgq2x7y

Parallel ABM for Electricity Distribution Grids: A Case Study [chapter]

Fanny Boulaire, Mark Utting, Robin Drogemuller
2014 Lecture Notes in Computer Science  
A fine-grained shared memory parallel implementation is presented, detailing the way the agents are grouped and executed on a multi-threaded machine, as well as the way the model is built (in a composable  ...  Current results show a medium level speedup of 2.6, but improvements are expected by incorporating newer distributed or parallel ABM schedulers into this implementation.  ...  The outputs of the simulation are identical whether the simulation is run in parallel or sequentially which is an important feature.  ... 
doi:10.1007/978-3-642-54420-0_55 fatcat:j5yw64bu2rh57d6ksjsidd3rry

Fast Crown Scheduling Heuristics for Energy-Efficient Mapping and Scaling of Moldable Streaming Tasks on Manycore Systems

Nicolas Melot, Christoph Kessler, Jörg Keller, Patrick Eitschberger
2015 ACM Transactions on Architecture and Code Optimization (TACO)  
We first present optimal off-line algorithms for separate and integrated crown scheduling based on integer linear programming (ILP). We make no restricting assumption about speedup behavior.  ...  We investigate the problem of generating energy-optimal code for a collection of streaming tasks that include parallelizable or moldable tasks on a generic manycore processor with dynamic discrete frequency  ...  We assume that our underlying machine consists of p identical processors, which can be frequency-scaled independently. We consider discrete frequency levels.  ... 
doi:10.1145/2687653 fatcat:ui46eht4wbdwbb72ariyk4rhkm

Fast Crown Scheduling Heuristics for Energy-Efficient Mapping and Scaling of Moldable Streaming Tasks on Many-Core Systems

Nicolas Melot, Christoph Kessler, Jörg Keller, Patrick Eitschberger
2015 Proceedings of the 18th International Workshop on Software and Compilers for Embedded Systems - SCOPES '15  
We first present optimal off-line algorithms for separate and integrated crown scheduling based on integer linear programming (ILP). We make no restricting assumption about speedup behavior.  ...  We investigate the problem of generating energy-optimal code for a collection of streaming tasks that include parallelizable or moldable tasks on a generic manycore processor with dynamic discrete frequency  ...  We assume that our underlying machine consists of p identical processors, which can be frequency-scaled independently. We consider discrete frequency levels.  ... 
doi:10.1145/2764967.2764975 dblp:conf/scopes/MelotK0E15 fatcat:u3gxpectozf6plldad75whvgua

A survey of pipelined workflow scheduling

Anne Benoit, Ümit V. Çatalyürek, Yves Robert, Erik Saule
2013 ACM Computing Surveys  
Efficient execution of such applications necessitates intelligent distribution of the application components and tasks on a parallel machine, and the execution can be orchestrated by utilizing task-, data  ...  A large class of applications need to execute the same workflow on different data sets of identical size.  ...  This formulation corresponds exactly to the problem of scheduling independent tasks on identical processors to minimize the makespan, that has been studied for a long time [Gra66; Gra69], and considered  ... 
doi:10.1145/2501654.2501664 fatcat:avzu434g2zbptm532lii42bqx4

Communication-Aware Load Balancing of the LU Factorization over Heterogeneous Clusters

Lucas Leandro Nesi, Lucas Mello Schnorr, Arnaud Legrand
2020 2020 IEEE 26th International Conference on Parallel and Distributed Systems (ICPADS)  
In this article, we build upon task-based runtimes' flexibility to study the interplay between static communicationaware data distribution strategies and dynamic scheduling of the linear algebra LU factorization  ...  First, to use fewer computing nodes towards the end to better match performance bounds and save computing power.  ...  StarPU uses the Sequential Task Flow (STF) paradigm [18] for task submission, where the application sequentially submits the tasks to the runtime that is responsible for scheduling then.  ... 
doi:10.1109/icpads51040.2020.00017 fatcat:ml3jhdap5jh5bifbpnjgispfh4

Resource-Aware Task Scheduling

Martin Tillenius, Elisabeth Larsson, Rosa M. Badia, Xavier Martorell
2015 ACM Transactions on Embedded Computing Systems  
Efficient utilization of these systems (in terms of resources as well as power) requires application software to be parallel, unless it can be assured that enough processes to occupy all cores are normally  ...  Today most computer systems, from embedded systems, via laptop and desktop computers, to high performance computer systems, are based on multicore architectures.  ...  Acknowledgment Parts of the experiments were performed on the Kalkyl cluster provided by SNIC through UPPMAX under project p2009014.  ... 
doi:10.1145/2638554 fatcat:tfzqjww7z5hjxoauosxqoocg64

Chance-Constrained Outage Scheduling using a Machine Learning Proxy [article]

Gal Dalal, Elad Gilboa, Shie Mannor, Louis Wehenkel
2018 arXiv   pre-print
To tackle tractability issues arising in large networks, we use machine learning to build a proxy for predicting outcomes of power system operation processes in this context.  ...  We propose a distributed scenario-based chance-constrained optimization formulation for this problem.  ...  We harness the power of machine learning and distributed computing to tractably perform multiple schedule assessments in parallel.  ... 
arXiv:1801.00500v1 fatcat:xid4ljfilzdsjf3kbn2vsc5uyu

A Parallel Framework for Parametric Maximum Flow Problems in Image Segmentation [article]

Vlad Olaru, Mihai Florea, Cristian Sminchisescu
2015 arXiv   pre-print
We present the case study of a state-of-the-art image segmentation algorithm based on graph cuts, Constrained Parametric Min-Cut (CPMC), that uses the parallel framework to solve parametric maximum flow  ...  The framework can also be used for performance evaluation of parallel implementations of maximum flow algorithms.  ...  To match the simplest paradigm of parallel, nonpreemptive scheduling for makespan minimization, namely that of Parallel and Identical Machines [19] , we gathered the running times of our one seed supergraph  ... 
arXiv:1509.06004v2 fatcat:nizlhk4pubbf3pbqooqoki2ftq

Execution Time Estimation for Workflow Scheduling

Artem M. Chirkin, A. S. Z. Belloum, Sergey V. Kovalchuk, Marc X. Makkes
2014 2014 9th Workshop on Workflows in Support of Large-Scale Science  
Estimation of the execution time is an important part of the workflow scheduling problem.  ...  The proposed estimation algorithm can be integrated easily into a wide class of schedulers as a separate module.  ...  ACKNOWLEDGMENT This work was partially supported by Government of Russian Federation, Grant 074-U01, project "Big data management for computationally intensive applications" (project #14613), and the Dutch  ... 
doi:10.1109/works.2014.11 dblp:conf/sc/ChirkinBKM14 fatcat:q4akrmt3fjespdnamgelnn4ls4

High-throughput bayesian computing machine with reconfigurable hardware

Mingjie Lin, Ilia Lebedev, John Wawrzynek
2010 Proceedings of the 18th annual ACM/SIGDA international symposium on Field programmable gate arrays - FPGA '10  
A Bayesian computing machine with 16 processing nodes was implemented with a Virtex-5 FPGA (XCV5LX155T-2) on a BEE3 (Berkeley Emulation Engine) platform.  ...  For a wide variety of sample Bayesian problems, comparing running the same network evaluation algorithm on a 2.4 GHz Core 2 Duo Intel processor and a GeForce 9400m using the CUDA software package, the  ...  The authors would like to thank Greg Gibeling and the Berkeley GateLib project for implementations of various HDL modules.  ... 
doi:10.1145/1723112.1723127 dblp:conf/fpga/LinLW10 fatcat:ejmbdbecg5hjzo27j3vdn3jstq

Designing Computational Clusters for Performance and Power [chapter]

Kirk W. Cameron, Rong Ge, Xizhou Feng
2007 Advances in Computers  
We describe power-aware and low power techniques to reduce the power profiles of parallel applications and mitigate the impact on performance.  ...  Yet, the demand for more powerful machines continues to grow. In this chapter, we motivate the need to reconsider the traditional performance-at-any-cost cluster design approach.  ...  Since all slave nodes are identical (as they should be and we experimentally confirmed), we use the M independent measurements on one node to emulate one measurement on M nodes.  ... 
doi:10.1016/s0065-2458(06)69002-5 fatcat:6piim42jtzcttnogp5gf45sz5e

Chance-Constrained Outage Scheduling using a Machine Learning Proxy

Gal Dalal, Elad Gilboa, Shie Mannor, Louis Wehenkel
2019 IEEE Transactions on Power Systems  
To tackle tractability issues arising in large networks, we use machine learning to build a proxy for predicting outcomes of power system operation processes in this context.  ...  SUBMITTED TO IEEE TRANSACTIONS ON POWER SYSTEMS 2 the coordination with day(s)-ahead and (intra)hourly operation.  ...  We harness the power of machine learning and distributed computing to tractably perform multiple schedule assessments in parallel.  ... 
doi:10.1109/tpwrs.2018.2889237 fatcat:7si6ehibxzb7reo66jrsistd3m

Simple Recurrent Units for Highly Parallelizable Recurrence

Tao Lei, Yu Zhang, Sida I. Wang, Hui Dai, Yoav Artzi
2018 Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing  
We demonstrate the effectiveness of SRU on multiple NLP tasks.  ...  ., 2017) on translation by incorporating SRU into the architecture. 1  ...  Acknowledgement We thank Alexander Rush and Yoon Kim for help with machine translation experiments, and Danqi Chen for help with SQuAD experiments.  ... 
doi:10.18653/v1/d18-1477 dblp:conf/emnlp/LeiZWDA18 fatcat:55hgrm6vjjbejanbzk5o435bke
« Previous Showing results 1 — 15 out of 6,027 results