Chao-Tung Yang's Articles: (5)
AbstractNowadays, NVIDIA's CUDA is a general purpose scalable parallel programming model for writing highly parallel applications. It provides several key abstractions – a hierarchy of thread blocks, shared memory, and barrier synchronization. This model has proven quite successful at programming multithreaded many core GPUs and scales transparently to hundreds of cores: scientists throughout industry and academia are already using CUDA to achieve dramatic speedups on production and research codes. In this paper, we propose a parallel programming approach using hybrid CUDA OpenMP, and MPI programming, which partition loop iterations according to the number of C1060 GPU nodes in a GPU cluster which consists of one C1060 and one S1070. Loop iterations assigned to one MPI process are processed in parallel by CUDA run by the processor cores in the same computational node.
AbstractIn this work, we propose a Grid-based intrusion detection platform, named Enhanced Dynamic Grid Intrusion Detection Environment (E-DGIDE), which is an extension of our previous system, DGIDE. The DGIDE exploits a Grid’s dynamic and abundant computing resources to detect intrusion packets. The E-DGIDE is a fault-tolerant platform that provides three types of standby mechanisms to prevent itself from crashing. The first two types are hot standby in which the standby subsystem performs the same task as its working subsystem. When the working subsystem fails, the standby takes over the current task immediately with less delay and without information passing. The other is cold standby. When the working subsystem cannot work properly, the E-DGIDE notifies the standby subsystem to take over. With these mechanisms, the reliability of an ordinary security system can be improved.
Highlights•This work proposes a novel method for managing green power of a virtual machine cluster in cloud computing environments.•A green power management scheme is proposed to determine how many physical machines should be run or turned off based on the gross occupied resource weight ratio of the virtual machine cluster.•When the gross occupied resource weight ratio is greater than a maximum tolerant occupied resource weight ratio, a standby physical machine in the non-running physical machines is selected and waken up to join as one of the running physical machines.•A resource allocation process is also used to distribute loads of the running physical machines such that the total number of the running physical machines can be flexibly dispatched to achieve the objective of green power management.
Highlights•The motivation of this paper is to attempt to resolve the problems of storing and sharing electronic medical records and medical images between different hospitals.•Specifically, this study develops a Medical Image File Accessing System (MIFAS) based on HDFS of Hadoop in cloud.•The proposed system can improve medical imaging storage, transmission stability, and reliability while providing an easy-to-operate management interface.•This paper focuses on the cloud storage virtualization technology to achieve high-availability services.•The experimental results show that the high reliability data storage clustering and fault tolerance capabilities can be achieved.
AbstractGrid computing environments with abundant resources can support innovative e-Learning applications, and are promising platforms for e-Learning. To support individualized and adaptive learning, teachers are encouraged to develop various teaching materials according to different requirements. However, traditional methodologies for designing teaching materials are time-consuming. To speed up the development process of teaching materials, our idea is to use a rapid prototyping approach which is based on automatic draft generation and Wiki-based revision. This paper presents the approach named WARP (Wiki-based Authoring by Rapid Prototyping), which is composed of five phases: (1) requirement verification, (2) query expansion, (3) teaching-material retrieval, (4) draft generation and (5) Wiki-based revision. A prototype system was implemented in grid environments. The evaluation was conducted using a two-group t-test design. Experimental results indicate that teaching materials can be rapidly generated with the proposed approach.