One among the most influential and popular data mining methods is the k-Means algorithm for cluster analysis. Techniques for improving the efficiency of k-Means have been largely explored in two main directions. The a...
详细信息
The lack of license management schemes in distributed environments is becoming a major obstacle for the commercial adoption of Grid or Cloud infrastructures. In this paper, we present a complete license management arc...
详细信息
Previous work on the use of the Teleo-Reactive technique in high level software development has shown it to be a viable approach for autonomic systems. A T-R program can recover from unexpected events without knowing ...
详细信息
We present a monitoring system for large-scale parallel and distributedcomputing environments that allows to trade-off accuracy in a tunable fashion to gain scalability without compromising fidelity. The approach rel...
详细信息
parallelcomputing is still evolving. Although there are great improvements in the field, still remains to be improved. The major research trend of parallelcomputing is now shifting to software technology from hardwa...
详细信息
This paper investigates a fast parallelcomputing scheme for the leaning control of a class of two-layered Networked Learning Control systems (NLCSs). This class of systems is subject to imperfect Quality of Service (...
详细信息
ISBN:
(纸本)9783642156205
This paper investigates a fast parallelcomputing scheme for the leaning control of a class of two-layered Networked Learning Control systems (NLCSs). This class of systems is subject to imperfect Quality of Service (QoS) in signal transmission, and requires a real-time fast learning. A parallel computational model for this task is established in the paper. Based on some of grid computing technologies and optimal scheduling, an effective scheme is developed to make full use of distributedcomputing resources, and thus to achieve a fast multi-objective optimization for the learning task under study. Experiments of the scheme show that it indeed provides a required fast on-line learning for NLCSs.
The Interconnection networks are essential elements in current computingsystems. For this reason, achieving the best network performance, even in congestion situations, has been a primary goal in recent years. In tha...
详细信息
Heterogeneous computing combines general purpose CPUs with accelerators to efficiently execute both sequential control-intensive and data-parallel phases of applications. Existing programming models for heterogeneous ...
详细信息
ISBN:
(纸本)9781605588391
Heterogeneous computing combines general purpose CPUs with accelerators to efficiently execute both sequential control-intensive and data-parallel phases of applications. Existing programming models for heterogeneous computing rely on programmers to explicitly manage data transfers between the CPU system memory and accelerator memory. This paper presents a new programming model for heterogeneous computing, called Asymmetric distributed Shared Memory (ADSM), that maintains a shared logical memory space for CPUs to access objects in the accelerator physical memory but not vice versa. The asymmetry allows light-weight implementations that avoid common pitfalls of symmetrical distributed shared memory systems. ADSM allows programmers to assign data objects to performance critical methods. When a method is selected for accelerator execution, its associated data objects are allocated within the shared logical memory space, which is hosted in the accelerator physical memory and transparently accessible by the methods executed on CPUs. We argue that ADSM reduces programming efforts for heterogeneous computingsystems and enhances application portability. We present a software implementation of ADSM, called GMAC, on top of CUDA in a GNU/Linux environment. We show that applications written in ADSM and running on top of GMAC achieve performance comparable to their counterparts using programmer-managed data transfers. This paper presents the GMAC system and evaluates different design choices. We further suggest additional architectural support that will likely allow GMAC to achieve higher application performance than the current CUDA model.
In grid and parallelsystems, backfilling has proven to be a very efficient method when parallel jobs are considered. However, it requires a job's runtime to be known in advance, which is not realistic with curren...
详细信息
ISBN:
(纸本)9780769539676
In grid and parallelsystems, backfilling has proven to be a very efficient method when parallel jobs are considered. However, it requires a job's runtime to be known in advance, which is not realistic with current technology. Various prediction methods do exist, but none is very accurate. In this study, we examine a grid system where both parallel and sequential jobs require service. Backfilling is used, but an error margin is added to a job's runtime prediction. The impact on system performance is examined and the results are compared with the optimal case of runtimes being known. Two different scheduling techniques are considered and a simulation model is used to evaluate system performance.
暂无评论