• ETRI Journal
• /
• v.45 no.6
• /
• pp.1035-1045
• /
• 2023

We propose a novel graphics processing unit (GPU) algorithm that can handle a large-scale 3D fast Fourier transform (i.e., 3D-FFT) problem whose data size is larger than the GPU's memory. A 1D FFT-based 3D-FFT computational approach is used to solve the limited device memory issue. Moreover, to reduce the communication overhead between the CPU and GPU, we propose a 3D data-transposition method that converts the target 1D vector into a contiguous memory layout and improves data transfer efficiency. The transposed data are communicated between the host and device memories efficiently through the pinned buffer and multiple streams. We apply our method to various large-scale benchmarks and compare its performance with the state-of-the-art multicore CPU FFT library (i.e., fastest Fourier transform in the West [FFTW]) and a prior GPU-based 3D-FFT algorithm. Our method achieves a higher performance (up to 2.89 times) than FFTW; it yields more performance gaps as the data size increases. The performance of the prior GPU algorithm decreases considerably in massive-scale problems, whereas our method's performance is stable.

• Current Research on Agriculture and Life Sciences
• /
• v.14
• /
• pp.37-47
• /
• 1996

The main objectives of this studies are to present the most desirable rice processing complex model system in a given our situations by comparision and analyzing the major factors and, also recommend the future prospect of the rice processing complex in Korea. There are 3 different rice processing complex models in Korea. Those are concrete bin, flat type steel bin and square bin. These systems have a lot of differences and have their own characteristics such as capital requirement, efficiency, storage capacity and quality controls. The major problems of the existing rice processing centers in Korea are high fixed cost and the unbalnced systems. Following is summary to solve this problems: 1. Development of the large scale harvester and high speed continuous dryer. 2. Quality inspective system of bulk grain and large scale temporary storage facilities. 3. Large size readjustment of arable land. 4. Select the convenient location of rice processing center and formulation of well equipment facilities.

• International Journal of CAD/CAM
• /
• v.9 no.1
• /
• pp.111-120
• /
• 2010

In this paper, we propose a volume rendering method using grid computing for large-scale volume data. Grid computing is attractive because medical institutions and research facilities often have a large number of idle computers. A large-scale volume data is divided into sub-volumes and the sub-volumes are rendered using grid computing. When using grid computing, different computers rarely have the same processor speeds. Thus the return order of results rarely matches the sending order. However order is vital when combining results to create a final image. Job-Scheduling is important in grid computing for volume rendering, so we use an obstacle-flag which changes priorities dynamically to manage sub-volume results. Obstacle-Flags manage visibility of each sub-volume when line of sight from the view point is obscured by other subvolumes. The proposed Dynamic Job-Scheduling based on visibility substantially increases efficiency. Our Dynamic Job-Scheduling method was implemented on our university's campus grid and we conducted comparative experiments, which showed that the proposed method provides significant improvements in efficiency for large-scale volume rendering.

• Journal of the Society of Naval Architects of Korea
• /
• v.53 no.3
• /
• pp.171-179
• /
• 2016

In order to study correlation of the propeller cavitation performance between a full-scale ship and a model ship for the MR Tanker, the full-scale ship and the model tests were conducted. The full-scale ship test is composed of cavitation observation, pressure fluctuation and noise measurements, which are conducted using 2 observation windows and 8 pressure transducers installed inside the full-scale ship above the propeller. The model test in the Large Cavitation Tunnel(LCT) was conducted at the same conditions as that of the full-scale ship and its results are compared with those of the full-scale ship. Through the model-ship correlation analysis, it is considered that the experimental technique for the MR Tanker class ship was verified in LCT.

• Journal of Radiation Protection and Research
• /
• v.40 no.1
• /
• pp.1-9
• /
• 2015

Compton cameras overcome several limitations of conventional mechanical collimation based gamma imaging devices, such as pin-hole imaging devices, due to its electronic collimation based on coincidence logic. Especially large-scale Compton camera has wide field of view and high imaging sensitivity. Those merits suggest that a large-scale Compton camera might be applicable to monitoring nuclear materials in large facilities without necessity of portability. To that end, our research group have made an effort to design a large-scale Compton camera for safeguard application. Energy resolution or position resolution of large-area detectors vary with configuration style of the detectors. Those performances directly affect the image quality of the large-scale Compton camera. In the present study, a series of Geant4 Monte Carlo simulations were performed in order to examine the effect of those detector parameters. Performance of the designed large-scale Compton camera was also estimated for various monitoring condition with realistic modeling. The conclusion of the present study indicates that the energy resolution of the component detector is the limiting factor of imaging resolution rather than the position resolution. Also, the designed large-scale Compton camera provides the 16.3 cm image resolution in full width at half maximum (angular resolution: $9.26^{\circ}$) for the depleted uranium source considered in this study located at the 1 m from the system when the component detectors have 10% energy resolution and 7 mm position resolution.

• KEPCO Journal on Electric Power and Energy
• /
• v.3 no.1
• /
• pp.17-21
• /
• 2017

Recently, the Korea Enhanced Power system Simulator (KEPS) was upgraded with new processor cards and various peripheral devices. The purpose of the upgrade is for the large-scale AC/DC simulation studies for different purposes such as FACTS, PMU applications, and SPS. The need for such study is now growing in KEPCO network. Frequently, traditional ways of performing the necessary study had been found to be less adequate for guiding the key decisions in the company. Therefore, the growing needs, as well as the revealed inadequacy of the traditional system studies, had been attributing to the momentum for the upgrade project. This paper explains the details of the upgrade project. As an example of presenting the effectiveness of the upgraded RTDS system in KEPRI-KEPCO, a large-scale AC/DC real time simulation case which includes the entire Korean network and a planned MMC-HVDC system is introduced.

• Proceedings of the Korean Geotechical Society Conference
• /
• 2005.03a
• /
• pp.627-632
• /
• 2005

In this study, a series of large-scale oedometer tests was performed to investigate the compressibility of rock fill materials. The testing samples were prepared to have three different grain size distributions and for each distribution, exist in two different states(dried and saturated). The test results indicated that particle breakages occurred mainly for the particles larger than 4.75mm in size and increased with increasing grain sizes. Also, it was found that, for a dry sample as it became well-graged, its compressibility decreased and accordingly, its tangent constrained modulus increased. A comparion between the samples in dry and saturated states revealed that compressibility of the materials increases with increasing water content. The values of tangent constrained modulus calculated for the tested dry samples were larger by about 10 to 20%, on average, than those for the saturated samples.

• Biomolecules & Therapeutics
• /
• v.16 no.1
• /
• pp.28-33
• /
• 2008

Atorvastatin calcium salt (1) was obtained through the preparation of lactone compound (8) from 2-((4R,6R)-6-(2-(2-(4-fluorophenyl)-5-isopropyl-3-phenyl-4-(phenylcarbamoyl)-1H-pyrrol-1-yl)-ethyl)-2-phenyl-1,3,2-dioxaborinan-4-yl)acetic acid tert-butyl ester (9) by hydrolysis in basic condition. Efficient hydrolysis of boronate compound 9 aimed at the viable synthesis for commercial production and purification of Atorvastatin calcium is reported. Detail studies of evaluation procedure are also reported.

• Animal cells and systems
• /
• v.3 no.1
• /
• pp.73-78
• /
• 1999

Although there are several methods for the preparation of mitochondrial DNA (mtDNA) from mammalian tissues, most are relatively long ultracentrifugation or manipulations by a small-scale method. We escribed a rapid method for large-scale extraction of mtDNA from human placental and horse liver tissues. The method is based on the preparation and homogenization of tissues, urification of crude mitochondria by differential centrifugations and isolation of mtDNA by alkaline Iysis. It was improved from Pre-existing methods by replacing some steps with simpler ones and discarding many others. This method gives a high yield of pure mtDNA(approximately 1-5mg from one placenta; ca. 400-600 g wet weight), depending on its sources (fresh tissue gave better results than frozen one). The resulting mtDNA indicated that this method can yield mtDNA in sufficient purity and quantity to identify the direct restriction analysis on agarose gel, random-primed labeling as a probe, and end labeling. Therefore, the method is ideal for obtaining good mtDNA samples to conduct routine restriction fragment length polymorphism (RFLP) analyses of natural populations for genetic studies.

• Journal of Information Technology Applications and Management
• /
• v.19 no.2
• /
• pp.249-263
• /
• 2012

Recently the open innovation is considered as one of the important issues for the survival strategy of a company in the industry. There have been several researches and practices on this topic. However, previous researches are mainly focused on the study of successful cases and some statistical analyses on the relationship between the open innovation and its performance. There are rare researches on the hindrance factors in the practical level. This paper tries to find some hinderance factors for the open innovation in Korean large-scale companies based on the WFGM(Want-Find-Get-Manage) model proposed by Gene Slowinski(2006). Especially this research focused on each processes of open innovation by in-depth interviews for two representative large companies in Korea. From the result of this research we come to know that the critical hindrance factor in the WANT step is the uncertain definition of the necessary technologies for the company. Issues on establishing and maintaining the innovation network are in the FIND step. Technology evaluation, technology introduction channel and technology copying issues are in the GET step. Communication issues, technology feasibility and lack of competition are in the MANAGE step. This paper finds some hindrance factors in each process step of open innovation, which gives some implications for the companies that want to adopt open innovation.