• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.12B
• /
• pp.1450-1458
• /
• 2011

The HPcN (Hybrid & high Performance Convergence Network) in research networks means environment which can provide both computing resource such as supercomputer, cluster and network resource to application researchers in the field of medical, bio, aerospace and e-science. The most representative research network in Korea, KREONET has been developing following technologies through the HERO (Hybrid Networking project for research oriented infrastructure) from 200S. First, we have constructed and deployed a control plane technology which can provide a connection oriented network dynamically. Second, the integrated resource management system technology has been developing for reservation and allocation of both computing and network resources, whenever users want to utilize them. In this paper, a testbed network is presented, which is possible to reserve and allocate network resource using the integrated resource management system. We reserve network resource through GNSI (Grid Network Service Interface) messages between GRS (Global Resource Scheduler) and NRM (Network Resource Manager) and allocate network resource through GUNI (Grid User Network Interface) messages between the NRM (network resource manager) and routers, based on reservation information provided from a user on the web portal. It is confirmed that GUNI interface messages are delivered from the NRM to each router at the starting of reservation time and traffic is transmitted through LSP allocated by the NRM.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.10
• /
• pp.250-258
• /
• 2019

Due to the problems associated with the aging of industrial complexes, research on the decline of industrial complexes is being conducted. In the case of decline, it is necessary to not respond immediately, but with a crisis, it is necessary to minimize the impact on the industrial complex through preemptive responses to the external environment and internal changes. Therefore, it is necessary to develop a crisis index that can systematically predict and evaluate changes in the industrial complex. In this research, a method for extracting the crisis index of an industrial complex is developed. We derive performance measures for developing the crisis index, deriving the relative importance of performance measures based on the analytical hierarchy process. Because units of performance measurement are different, a normalization method is developed to sensitively reflect change. Based on the relative importance and normalized values of the performance measures, the crisis index of the industrial complex is developed and applied to a national industrial complex in order to verify its applicability.

• KIPS Transactions on Computer and Communication Systems
• /
• v.11 no.10
• /
• pp.333-344
• /
• 2022

In a container-based cloud environment, multiple containers can share a graphical processing unit (GPU), and GPU sharing can minimize idle time of GPU resources and improve resource utilization. However, in a cloud environment, GPUs, unlike CPU or memory, cannot logically multiplex computing resources to provide users with some of the resources in an isolated form. In addition, containers occupy GPU resources only when performing GPU operations, and resource usage is also unknown because the timing or size of each container's GPU operations is not known in advance. Containers unrestricted use of GPU resources at any given point in time makes managing resource contention very difficult owing to where multiple containers run GPU tasks simultaneously, and GPU tasks are handled in black box form inside the GPU. In this paper, we propose a container management technique to prevent performance degradation caused by resource competition when multiple containers execute GPU tasks simultaneously. Also, this paper demonstrates the efficiency of container management techniques that analyze and propose the problem of degradation due to resource competition when multiple containers execute GPU tasks simultaneously through experiments.

• International Journal of Highway Engineering
• /
• v.20 no.2
• /
• pp.19-25
• /
• 2018

PURPOSES : This purpose of this study is to analyze the effect to autogenous shrinkage of the top-layer material of a two-lift concrete pavement mixing both silica fume and polymer powder. METHODS : The bottom-layer of a two-lift concrete pavement was paved with original portland cement (OPC) with a 20~23 cm thickness. Additionally, the top-layer which is directly exposed to the environment and vehicles was paved with a high-performance concrete (HPC) with a 7~10 cm thickness. These types of pavements can achieve a long service life by reducing joint damage and increasing the abrasion and scaling resistance. In order to integrate the different bottom and top layer materials, autogenous shrinkage tests were performed in this study according to the mixing ratio of silica fume and polymer powder, which are the admixture of the top-layer material. RESULTS : Autogenous shrinkage decreased when polymer powder was used in the mix. Contrary to this, autogenous shrinkage tended to rise with increasing silica fume content. However, the effects were not significant when small amounts of polymer powder were used (3% and 11%). CONCLUSIONS : The durability and compressive strength increase when silica fume is used in the mix. The flexural strength considerably increases and autogenous shrinkage of concrete decreases when polymer powder is used in the mix. As seen from above, the proper use of these materials improves not only durability, but also autogenous shrinkage, leading to better shrinkage crack control in the concrete.

• Journal of the Korean Society for Marine Environment & Energy
• /
• v.19 no.3
• /
• pp.218-226
• /
• 2016

Field observations were used to study the characteristics and influence of groundwater level fluctuations on vegetation development on the natural beach of a sandy barrier island, in the Nakdong River estuary. The spatial/temporal fluctuations of the groundwater level and the interactions with the external forces (weather, ocean wave and tide) were analyzed. The results indicated that when it rains the groundwater level rises. During summer, when precipitation intensity is greater than 20 mm/hour, it rose rapidly over 20 cm. Subsequently, it fell gradually during periods of no precipitation. Seasonal characteristics indicated that the groundwater level was high during the summer rainy season and tended to fall in the winter dry season. The time-averaged groundwater level, observed from the four observations over 3 years (2012-2014), was about 1.47 m, higher than mean sea level (M.S.L.). It was shown that the average annual groundwater level rises toward the land rather than showing intertidal patterns observation. Differences in the presence or absence of a coastal sand dunes affected the progress of vegetation. In other words, in environments of saltwater intrusion where the groundwater level varies, dependent on the distance from the shoreline and bottom slope, sand dunes can be provided to affect soil conditions and groundwater, so that vegetation can be grown reliably.

• KIPS Transactions on Software and Data Engineering
• /
• v.2 no.2
• /
• pp.81-90
• /
• 2013

During the past decade, many changes and attempts have been tried and are continued developing new technologies in the computing area. The brick wall in computing area, especially power wall, changes computing paradigm from computing hardwares including processor and system architecture to programming environment and application usage. The high performance computing (HPC) area, especially, has been experienced catastrophic changes, and it is now considered as a key to the national competitiveness. In the late 2000's, many leading countries rushed to develop Exascale supercomputing systems, and as a results tens of PetaFLOPS system are prevalent now. In Korea, ICT is well developed and Korea is considered as a one of leading countries in the world, but not for supercomputing area. In this paper, we describe architecture design of MAHA supercomputing system which is aimed to develop 300 TeraFLOPS system for bio-informatics applications like human genome analysis and protein-protein docking. MAHA supercomputing system is consists of four major parts - computing hardware, file system, system software and bio-applications. MAHA supercomputing system is designed to utilize heterogeneous computing accelerators (co-processors like GPGPUs and MICs) to get more performance/$, performance/area, and performance/power. To provide high speed data movement and large capacity, MAHA file system is designed to have asymmetric cluster architecture, and consists of metadata server, data server, and client file system on top of SSD and MAID storage servers. MAHA system softwares are designed to provide user-friendliness and easy-to-use based on integrated system management component - like Bio Workflow management, Integrated Cluster management and Heterogeneous Resource management. MAHA supercomputing system was first installed in Dec., 2011. The theoretical performance of MAHA system was 50 TeraFLOPS and measured performance of 30.3 TeraFLOPS with 32 computing nodes. MAHA system will be upgraded to have 100 TeraFLOPS performance at Jan., 2013.

• KIPS Transactions on Computer and Communication Systems
• /
• v.11 no.11
• /
• pp.381-394
• /
• 2022

Recently, due to the personalization and customization of data, Internet-based services have increased requirements for real-time processing, such as real-time AI inference and data analysis, which must be handled immediately according to the user's situation or requirement. Real-time tasks have a set deadline from the start of each task to the return of the results, and the guarantee of the deadline is directly linked to the quality of the services. However, traditional container systems are limited in operating real-time tasks because they do not provide the ability to allocate and manage deadlines for tasks executed in containers. In addition, tasks such as AI inference and data analysis basically utilize graphical processing units (GPU), which typically have performance impacts on each other because performance isolation is not provided between containers. And the resource usage of the node alone cannot determine the deadline guarantee rate of each container or whether to deploy a new real-time container. In this paper, we propose a monitoring technique for tracking and managing the execution status of deadlines and real-time GPU tasks in containers to support real-time processing of GPU tasks running on containers, and a node list management technique for container placement on appropriate nodes to ensure deadlines. Furthermore, we demonstrate from experiments that the proposed technique has a very small impact on the system.