23,037 Hits in 9.4 sec

Adaptation Strategies in Multiprocessors System on Chip [chapter]

Remi Busseuil, Gabriel Marchesan Almeida, Luciano Ost, Sameer Varyani, Gilles Sassatelli, Michel Robert
2012 IFIP Advances in Information and Communication Technology  
Memory organization is a crucial criterion in MPSoC performance optimization, as memory access latency of remote data increases exponentially with respect to the number of cores.  ...  This chapter explors three different adaptable mechanisms, and shows their benefits : frequency scaling, task migration techniques and memory organization.  ...  of two task migration mechanisms into an RTL-modeled NoC-based homogenous MPSoC, leading to accurate results, and (iii) a hybrid memory architecture that can be employed to optimize the performance of  ... 
doi:10.1007/978-3-642-28566-0_10 fatcat:lqxg2ujikndgjjvnxrli6x5cza

Cluster-based Communication and Load Balancing for Simulations on Dynamically Adaptive Grids

Martin Schreiber, Hans-Joachim Bungartz
2014 Procedia Computer Science  
For data migration, the stack-based intra-cluster communication allows a very low memory footprint for data migration and the RLE leads to efficient updates of connectivity information.  ...  Furthermore, we demonstrate the applicability of cluster-based optimizations on distributed-memory systems.  ...  Acknowledgements We like to thank the Munich Centre of Advanced Computing for for funding this project by providing computing time on the MAC Cluster.  ... 
doi:10.1016/j.procs.2014.05.209 fatcat:iguhbpp4rnhnddvefqtod57tfi

ParadisEO: A Framework for the Reusable Design of Parallel and Distributed Metaheuristics

S. Cahon, N. Melab, E.-G. Talbi
2004 Journal of Heuristics  
ParadisEO provides a broad range of features including evolutionary algorithms (EA), local searches (LS), the most common parallel and distributed models and hybridization mechanisms, etc.  ...  Their implementation is portable on distributed-memory machines as well as on shared-memory multiprocessors, as it uses standard libraries such as MPI, PVM and PThreads.  ...  On shared memory multiprocessors the implementation is multi-threaded and the migrations are memory copy-driven.  ... 
doi:10.1023/ fatcat:72xaseacx5bqfgewn6puv533vm

ParaDisEO-Based Design of Parallel and Distributed Evolutionary Algorithms [chapter]

S. Cahon, N. Melab, E. -G. Talbi, M. Schoenauer
2004 Lecture Notes in Computer Science  
ParaDisEO is a framework dedicated to the design of parallel and distributed metaheuristics including local search methods and evolutionary algorithms. This paper focuses on the latter aspect.  ...  These models can be deployed on distributed memory machines as well as on shared memory multi-processors, taking advantage of the shared memory in the latter case.  ...  The extensions include local search methods (descent search, simulated annealing and tabu search for combinatorial optimization, gradient-based search for numerical optimization), hybridization mechanisms  ... 
doi:10.1007/978-3-540-24621-3_18 fatcat:c5lofyvntjdhblmztvgzp2a6we

Strategies for dynamic memory allocation in hybrid architectures

Peter Bertels, Wim Heirman, Dirk Stroobandt
2009 Proceedings of the 6th ACM conference on Computing frontiers - CF '09  
The proposed self-learning memory allocation strategy finds the optimal location for each Java object's data by means of runtime profiling.  ...  This article shows how we minimise this communication cost by dynamically chosing an optimal data layout in the Java heap memory which is distributed over both the accelerator and the processor memory.  ...  In our hybrid architecture this cost is caused by the nonuniformity of access times to the distributed heap memory, formed by main memory and local memory of the accelerator.  ... 
doi:10.1145/1531743.1531778 dblp:conf/cf/BertelsHS09 fatcat:5y4a7oe7anfsdc3uuj6cbv43jm

A Systematic Review of Existing VM Migration Techniques

Ramandeep Kaur, Supreet Kaur
2017 International Journal of Advanced Research in Computer Science and Software Engineering  
Jinhua Hu[2010][16] presented a scheduling strategy on load balancing of VM resources based on genetic algorithm.  ...  The model validated that there is a need of different statistical distribution to represent CPU, memory and total number of instructions.  ...  Hybrid Migration: Hybrid migration hybrids pre copy and post copy migration techniques .  ... 
doi:10.23956/ijarcsse/v7i5/0180 fatcat:nfv5d34vkrb2vdvflc2yjf7syy

Invasive Compute Balancing for Applications with Shared and Hybrid Parallelization

Martin Schreiber, Christoph Riesinger, Tobias Neckel, Hans-Joachim Bungartz, Alexander Breuer
2014 International journal of parallel programming  
We present a coredistribution scheduler which realizes the migration of computational power by distributing the cores depending on the requirements specified by one or more parallel program instances.  ...  The invasive paradigm using compute migration represents an efficient alternative to classical data migration approaches for such algorithms in HPC.  ...  Hybrid parallelization Invasive Computing for hybrid parallelized applications involves a mix of shared and distributed memory parallelization strategies.  ... 
doi:10.1007/s10766-014-0336-3 fatcat:3epov6cgnzeetpkhtyazctlrzy

Eagle Strategy with Cauchy Mutation Particle Swarm Optimization for Energy Management in Cloud Computing

Archana Kollu, Koneru Lakshmaiah Education Foundation, Sucharita Vadlamudi, Koneru Lakshmaiah Education Foundation
2020 International Journal of Intelligent Engineering and Systems  
We propose, Eagle Strategy (ES) based Modified Particle Swarm Optimization (ES-MPSO) to minimize the energy consumption and SLA violation.  ...  The Eagle Strategy method is applied due to its efficient local optimization technique.  ...  Author contributions The first author of the paper contributed in conceptualization, methodology, software validation and verification, formal analysis, investigation, resources, data curation, writing  ... 
doi:10.22266/ijies2020.1231.05 fatcat:4rax46syd5c5zluhnnbhkgblme

A survey on load balancing algorithms for virtual machines placement in cloud computing

Minxian Xu, Wenhong Tian, Rajkumar Buyya
2017 Concurrency and Computation  
A detailed classification targeting load balancing algorithms for VM placement in cloud data centers is investigated and the surveyed algorithms are classified according to the classification.  ...  The goal of this paper is to provide a comprehensive and comparative understanding of existing literature and aid researchers by providing an insight for potential future enhancements.  ...  with project ID 61672136 and 61650110513.  ... 
doi:10.1002/cpe.4123 fatcat:fzbfejblzrdyjhaezylkg4ye4q

Leveraging Transparent Data Distribution in OpenMP via User-Level Dynamic Page Migration [chapter]

Dimitrios S. Nikolopoulos, Theodore S. Papatheodorou, Constantine D. Polychronopoulos, Jesús Labarta, Eduard Ayguadé
2000 Lecture Notes in Computer Science  
The page migration engine improves transparently the locality of memory references at the page level on behalf of the application.  ...  The vehicle for implementing these facilities in OpenMP without modifying the programming model or exporting data distribution details to the programmer is user-level dynamic page migration [9,10].  ...  Plain shared-memory programming models hide the details of data distribution from the programmer and rely on the operating system for laying out the data in a locality-aware manner.  ... 
doi:10.1007/3-540-39999-2_40 fatcat:e7tdiomq3nc6bgemkkuufaacfi

Cost-Based Assessment of Partitioning Algorithms of Agent-Based Systems on Hybrid Cloud Environments [article]

Chahrazed Labba, Narjès Bellamine Ben Saoud
2017 arXiv   pre-print
Distributing agent-based simulators reveals many challenges while deploying them on a hybrid cloud infrastructure.  ...  The first experimental results show that, for a given agent-based model, a good partitioning method used with the suitable hybrid cloud environment lead to an efficient and economic deployment.  ...  To optimize the cost models associated with the use of hybrid clouds, CloudWard Bound implements a beneficial migration strategy based on considering the application characteristics including the workload  ... 
arXiv:1709.05708v1 fatcat:73g6gbtog5cn5jiy2rpduuvkv4

An Energy-Efficient and Fast Scheme for Hybrid Storage Class Memory in an AIoT Terminal System

Hao Sun, Lan Chen, Xiaoran Hao, Chenji Liu, Mao Ni
2020 Electronics  
In addition, a fast mode with a tmpfs-based, in-memory file system is applied to hybrid storage class memory to reduce the number of data movements between memory and external storage.  ...  This paper presents a hybrid storage class memory system to reduce the energy consumption and optimize IO performance.  ...  Acknowledgments: The authors would like to thank anonymous reviewers for their special effort. Conflicts of Interest: The authors declare no conflict of interest.  ... 
doi:10.3390/electronics9061013 fatcat:pzqb5c2mxjaolewur6ae57ti5u

Special section on advances in methods for adaptive multicore systems

Amir-Mohammad Rahmani, Pasi Liljeberg, Juha Plosila, Hannu Tenhunen
2014 Journal of Supercomputing  
Based on this, Holmbacka et al., in their article entitled "A Task Migration Mechanism for Distributed Many-Core Operating Systems" present a lightweight task migration mechanism explicitly for distributed  ...  They investigate the overhead of migrating tasks on a distributed OS running both on a bus based platform and a many-core networks-onchip. Multi-core computing has gone mobile.  ... 
doi:10.1007/s11227-014-1217-7 fatcat:pboxlhq5ajcx7jvb5li3f7c6va

A Comparison of Implementation Strategies for Nonuniform Data-Parallel Computations

Salvatore Orlando, Raffaele Perego
1998 Journal of Parallel and Distributed Computing  
Workload imbalance can be mitigated by cyclically distributing data and associated computations, or by employing adaptive strategies which build a more balanced schedule at run{time, on the basis of the  ...  W e fed the benchmark code with synthetic input data sets built on the basis of a load imbalance model and we report and compare the results obtained.  ...  The rst hybrid approach is based on the Inspector-Executor paradigm 2].  ... 
doi:10.1006/jpdc.1998.1456 fatcat:dpu3rbatdned7evd5pvvmmnsii

Geographical Locality and Dynamic Data Migration for OpenMP Implementations of Adaptive PDE Solvers [chapter]

Markus Nordén, Henrik Löf, Jarmo Rantakokko, Sverker Holmgren
2008 Lecture Notes in Computer Science  
of misplaced data, (3) that a migrate-on-next-touch directive works well whereas the first-touch strategy is less advantageous for programs exhibiting a dynamically changing memory access patterns, and  ...  On cc-NUMA multi-processors, the non-uniformity of main memory latencies motivates the need for co-location of threads and data. We call this special form of data locality, geographical locality.  ...  In a NUMA system, the latency for a main memory access depends on whether data is accessed at a local memory location or at a remote location.  ... 
doi:10.1007/978-3-540-68555-5_31 fatcat:5grtv5jzrzdkfpzbg3fghhitha
« Previous Showing results 1 — 15 out of 23,037 results