Complex networks are relational data sets commonly represented as graphs. The analysis of their intricate structure is relevant to many areas of science and commerce, and data sets may reach sizes that require distrib...
详细信息
ISBN:
(纸本)9781450338547
Complex networks are relational data sets commonly represented as graphs. The analysis of their intricate structure is relevant to many areas of science and commerce, and data sets may reach sizes that require distributed storage and processing. We describe and compare programming models for distributed computing with a focus on graph algorithms for large-scale complex network analysis. Four frameworks - GraphLab, Apache Giraph, Giraph++ and Apache Flink - are used to implement algorithms for the representative problems Connected Components, Community Detection, PageRank and Clustering Coefficients. The implementations are executed on a computer cluster to evaluate the frameworks' suitability in practice and to compare their performance to that of the single-machine, shared-memory parallel network analysis package NetworKit. Out of the distributed frameworks, GraphLab and Apache Giraph generally show the best performance. In our experiments a cluster of eight computers running Apache Giraph enables the analysis of a network with about 2 billion edges, which is too large for a single machine of the same type. However, for networks that fit into memory of one machine, the performance of the shared-memory parallel implementation is far better than the distributed ones. The study provides experimental evidence for selecting the appropriate framework depending on the task and data volume.
The proceedings contain 6 papers. The topics discussed include: tuning stationary iterative solvers for fault resilience;a scalable randomized least squares solver for dense over-determined systems;mixed-precision blo...
ISBN:
(纸本)9781450340113
The proceedings contain 6 papers. The topics discussed include: tuning stationary iterative solvers for fault resilience;a scalable randomized least squares solver for dense over-determined systems;mixed-precision block gram Schmidt orthogonalization;a parallel ensemble kalman filter implementation based on modified;weighted dynamic scheduling with many parallelism grains for offloading of numerical workloads to multiple varied accelerators;and on efficient Monte Carlo preconditioners and hybrid Monte Carlo methods.
The proceedings contain 9 papers. The topics discussed include: DeltaFS: exascale file systems scale better without dedicated servers;taming the cloud object storage with MOS;experiences in using OS-level virtualizati...
ISBN:
(纸本)9781450340083
The proceedings contain 9 papers. The topics discussed include: DeltaFS: exascale file systems scale better without dedicated servers;taming the cloud object storage with MOS;experiences in using OS-level virtualization for block I/O;BAD-Check: bulk asynchronous distributed checkpointing;tackling the reproducibility problem in storage systems research with declarative experiment specifications;comparative I/O workload characterization of two leadership class storage clusters;heavy-tailed distribution of parallel I/O system response;pattern-driven parallel I/O tuning;and automatic and transparent I/O optimization with storage integrated application runtime support.
Welcome to Cluj-Napoca, Romania, at the 2015 ieee 11th International Conference on Intelligent Computer Communication and Processing. The main goal of this conference is providing an international forum and discussion...
详细信息
Human brain communicates information by means of electro-chemical reactions and processes it in a parallel, distributed manner. Computational models of neurons at different levels of details are used in order to make ...
详细信息
ISBN:
(纸本)9781467366700
Human brain communicates information by means of electro-chemical reactions and processes it in a parallel, distributed manner. Computational models of neurons at different levels of details are used in order to make predictions for physiological dysfunctions. advances in the field of brain simulations and brain computer interfaces have increased the complexity of this modeling process. With a focus to build large-scale detailed networks, we used high performance computing techniques to model and simulate the granular layer of the cerebellum. Neuronal firing patterns of cerebellar granule neurons were modeled using two mathematical models Hodgkin- Huxley (HH) and Adaptive Exponential Leaky Integrate and Fire(AdEx). The performance efficiency of these modeled neurons was tested against a detailed multicompartmental model of the granule cell. We compared different schemes suitable for large scale simulations of cerebellar networks. Large networks of neurons were constructed and simulated. Graphic Processing Units (GPU) was employed in the pleasantly parallel implementation while Message Passing Interface (MPI) was used in the distributed computing approach. This allowed to explore constraints of different parallel architectures and to efficiently load balance the tasks by maximally utilizing the available resources. For small scale networks, the observed absolute speedup was 6X in an MPI based approach with 32 processors while GPUs gave 10X performance gain compared to a single CPU implementation. In large networks, GPUs gave approximately 5X performance gain in processing time compared to the MPI implementation. The results enabled us to choose parallelization schemes suitable for large-scale simulations of cerebellar circuits. We are currently extending the network model based on large scale simulations evaluated in this paper and using a hybrid - heterogeneous MPI based multi-GPU architecture for incorporating millions of cerebellar neurons for assessing physiologi
The proceedings contain 8 papers. The topics discussed include: a workflow runtime environment for Manycore parallel architectures;orchestrating workflows over heterogeneous networking infrastructures;towards efficien...
ISBN:
(纸本)9781450339896
The proceedings contain 8 papers. The topics discussed include: a workflow runtime environment for Manycore parallel architectures;orchestrating workflows over heterogeneous networking infrastructures;towards efficient scheduling of data intensive high energy physics workflows;contemporary challenges for data-intensive scientific workflow management systems;co-sites: the autonomous distributed dataflows in collaborative scientific discovery;inter-language parallel scripting for distributed-memory scientific computing;dynamically reconfigurable workflows for time-critical applications;enabling workflow repeatability with virtualization support;and workflow provenance: an analysis of long term storage costs.
This paper discusses an efficient parallel implementation of the ensemble Kalman filter based on the modified Cholesky decomposition. The proposed implementation starts with decomposing the domain into sub-domains. In...
详细信息
The proceedings contain 26 papers. The topics discussed include: response time schedulability analysis for hard real-time systems accounting DVFS latency on heterogeneous cluster-based platform;calculation of worst-ca...
ISBN:
(纸本)9781467394192
The proceedings contain 26 papers. The topics discussed include: response time schedulability analysis for hard real-time systems accounting DVFS latency on heterogeneous cluster-based platform;calculation of worst-case execution time for multicore processors using deterministic execution;a n unconventional computing technique for ultra- fast and ultra-low power data mining;dedicated network for distributed configuration in a mixed-signal wireless sensor node circuit;energy management via PI control for data parallel applications with throughput constraints;dynamic current reduction of CMOS digital circuits through design and process optimization;unified power format (UPF) methodology in a vendor independent flow;a synchronous sub-threshold ultra-low power processor;and constructing stability-based clock gating with hierarchical clustering.
Scripting languages such as Python and R have been widely adopted as tools for the development of scientific software because of the expressiveness of the languages and their available libraries. However, deploying sc...
详细信息
We introduce a new Manycore Workflow Runtime Environment (MWRE) to efficiently enact traditional scientific workflows on modern manycore computing architectures. In contrast to existing engines that enact workflows ac...
详细信息
暂无评论