• Proceedings of the Korea Technology Innovation Society Conference
• /
• 2014.10a
• /
• pp.59-66
• /
• 2014

• Proceedings of the Korea Technology Innovation Society Conference
• /
• 2014.10a
• /
• pp.603-612
• /
• 2014

• Proceedings of the Korea Technology Innovation Society Conference
• /
• 2015.11a
• /
• pp.832-837
• /
• 2015

• Journal of Scientific & Technological Knowledge Infrastructure
• /
• s.23
• /
• pp.14-17
• /
• 2006

• Journal of Scientific & Technological Knowledge Infrastructure
• /
• s.26
• /
• pp.8-11
• /
• 2007

• Journal of Digital Contents Society
• /
• v.16 no.5
• /
• pp.691-697
• /
• 2015

Supercomputers are used for many different areas including new product design of industries as well as state-of-the-art science and technology for large amount of computational needs. Tachyon is a 4th supercomputer built at KISTI that is a high-performance parallel computing system with 3,200 computing nodes and infrastructures. This system is currently about 10,000 users and over 170 organizations are used, the number of jobs they are performing work in batch type form through a scheduler. Also, this system logs lots of job scripts, execution environment, library, job status from the job submit to end. In this paper, we analyzed batch jobs information from Sun Grid Engine, that use as a scheduler in Tachyon system, and job executed information in Tachyon System. In particular, we distinguished the fail jobs from the all tasks that users perform and we analyzed the cause of failure. Among them, we can extracted some of jobs that can be regarded as normal jobs through the improvement in those works logged as all of fail jobs.

• Journal of the Korea Convergence Society
• /
• v.10 no.11
• /
• pp.1-7
• /
• 2019

Recently, various approaches to detect network attacks using machine learning have been studied and are being applied to detect new attacks and to increase precision. However, the machine learning method is dependent on feature extraction and takes a long time and complexity. It also has limitation of performace due to learning data imbalance. In this study, we propose a method to solve the degradation of classification performance due to imbalance of learning data among the limit points of detection system. To do this, we generate data using Generative Adversarial Networks (GANs) and propose a classification method using Convolutional Neural Networks (CNNs). Through this approach, we can confirm that the accuracy is improved when applied to the NSL-KDD and UNSW-NB15 datasets.

• KIPS Transactions on Computer and Communication Systems
• /
• v.12 no.7
• /
• pp.209-216
• /
• 2023

Modern HPC systems are equipped with many-core CPUs with dozens of cores. When performing parallel I/O in such a system, there is a limit to scalability due to the problem of the LRU lock management policy of the Linux system. The study proposes an improved FinerLRU to solve this problem. Our new FinerLRU improves the parallel write performance of file systems using the buffer cache through granular lock management by increasing the number of LRU locks upto the maximum number of cores. The proposed method was implemented in Linux 5.18.11, and the performance was measured on two types of CPUs, Intel Icelake Xeon and Intel Knights landing, with different characteristics, and it was found that a performance improvement of about two times can be obtained in both types of systems.

• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.9
• /
• pp.685-689
• /
• 2009

Flash memory has several features such as low~power consumption and fast access so that there has been various research on using flash memory as new storage. Especially the Solid State Disk which is composed of flash memory chips has recently replaced the hard disk. At present, SSD adopts the multi-channel and multi-way architecture to exploit advantages of parallel access. In this architecture, data are written on SSD in a unit of a superblock which is composed of multiple blocks in which some blocks are put together. This paper proposes two schemes of selecting, segmenting and re-composing victim superblocks to optimize concurrent processing when a buffer flush occurs. The experimental results show that 35% of superblock- based write operations is reduced by selecting victims and additional 9% by composition of superblock.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.17 no.11
• /
• pp.340-347
• /
• 2016

Since the 1970s, manufacturing has been one of the key driving forces that has led to Korea's economic growth. However, this growth rate has been reduced significantly since the 2000s, and shows that revenues and employment are steadily decreasing. In addition, while manufacturing investment in Korea has dropped sharply, the United States, Germany, Japan, and other major countries have increased investment in manufacturing. These countries have promoted manufacturing innovation strategies that include the convergence of information and communications technologies (ICT) and manufacturing. For manufacturing innovation, it is important for time and cost savings required for product development to be achieved by changes in the production process, especially product design. Modeling and simulation (M&S) is a process that replaces physical product design, mockup making, and testing, with virtual product creation (modeling) and engineering analysis (simulation). In this paper, we analyze the economic ripple effect of supercomputing M&S using an input-output model technique based on the input-output tables published by the Bank of Korea. When we set the M&S budget (about US$16 million for the last 10 years) of the Korea Institute of Science and Technology Information (KISTI) as input coefficients, the effect on production inducement, value-added inducement, and employment inducement was analyzed to be US$24 million, US$13.4 million, and 267, respectively.