• JSTS:Journal of Semiconductor Technology and Science
• /
• 제12권3호
• /
• pp.297-312
• /
• 2012

For future multi-core processors, 3D integration is regarded as one of the most promising techniques since it improves performance and reduces power consumption by decreasing global wire length. However, 3D integration causes serious thermal problems since the closer proximity of heat generating dies makes existing thermal hotspots more severe. Conventional air cooling schemes are not enough for 3D multi-core processors due to the limit of the heat dissipation capability. Without more efficient cooling methods such as liquid cooling, the performance of 3D multi-core processors should be degraded by dynamic thermal management. In this paper, we examine the architectural impact of cooling methods on the 3D multi-core processor to find potential benefits of liquid cooling. We first investigate the thermal behavior and compare the performance of two different cooling schemes. We also evaluate the leakage power consumption and lifetime reliability depending on the temperature in the 3D multi-core processor.

• 한국원자력학회:학술대회논문집
• /
• 한국원자력학회 1995년도 추계학술발표회논문집(1)
• /
• pp.41-47
• /
• 1995

A parametric study bas been performed for the various refueling schemes of CANDU 6 reactor loaded with reference DUPIC fuel. The optimum discharge burnup was determined such that the peak bundle power is minimized for the equilibrium core. Based on the results of instantaneous core calculation using patterned random age distributions, it was decided to perform the refueling simulations only for 2-bundle and 4-bundle shift refueling schemes. The 600 FPD simulation has shown that the operational margins of the channel and bundle power to the license limits are 7.9% and 17.1%, respectively, for 2-bundle shift refueling scheme. The 4-bundle shift refueling scheme also satisfies the license limits and the operational margins of the channel and bundle power are 7.1% and 9.8%, respectively. The result of refueling simulation indicate the possibility of using reference DUPIC fuel in current CANDU 6 reactor.

• Nuclear Engineering and Technology
• /
• 제29권4호
• /
• pp.299-309
• /
• 1997

The neutronic feasibility of typical Korean three-loop 900MWe class PWR core loaded with mixed oxide fuels for both annual and 18-month cycle strategies has been investigated as a means for spent fuel management. For this study, a method of determining equivalent plutonium content was developed under the equivalence concept which gives the same cycle length as uranium fuel. Optimal plutonium zoning within the MOX assembly was also designed with the aim of minimizing the peak md power. Conceptual core designs hate hen developed for equilibrium cycle with the following variations: annual and 18-month cycle, 1/3 and full MOX loading schemes, and typical and high moderation lattice. The analysis of key core physics parameters shows that in all cases considered satisfactory core designs seem to be feasible, though addition of control rod system and change in Technical Specification for soluble boron concentration are required for full MOX loading in order to meet the current design requirements.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2000년도 추계종합학술대회 논문집(3)
• /
• pp.9-12
• /
• 2000

Genetic algorithm is a searching method which based on the law of the survival of the fittest. Multi-population Genetic Algorithm is a modified form of Genetic Algorithm, which was devised for covering the defect of general genetic algorithm. The core of multi-population genetic algorithm is said to be the migration schemes. The fitness-based migration scheme and the random migration scheme are currently used. In this paper, a new migration scheme, ‘the migration scheme between groups’, is suggested, and compared to the general two migration schemes.

• Nuclear Engineering and Technology
• /
• 제51권2호
• /
• pp.377-383
• /
• 2019

Targeting at simulating the application of thorium-uranium (TU) fuel in the CANDU-6 reactor, this paper analyzes the process using the code DRAGON/DONJON where the discrete TU fuel pins are applied in the CANDU-6 reactor under the time-average equilibrium refueling. The results show that the coolant void reactivity of the assembly analyzed in this paper is lower than that of 37-element bundle cell with natural uranium and 37-element bundle cell with mixed TU fuel pins; that the max time-average channel/bundle power of the core meets the limits - less than 6700kW/860 kW; that the fuel conversion ratio is higher than that of the CANDU-6 reactor with natural uranium; and that the exit burnup increases to 13400 MWd/tU. Thus, the simulation in this paper with the fuel in the 37-element bundle cell using discrete TU fuel pins can be considered to be applied in CANDU-6 reactor with adequate modifications of the core structure and operating modes.

• 한국도서관정보학회지
• /
• 제21권
• /
• pp.195-229
• /
• 1994

This article investigates the normative principles suggested by Rangannathan as the guiding principles for his theories, consisting of basic laws, fundamental laws, canons, principles and postulates. His five basic laws and five laws of library science are re-interpreted from the view point of library classification. And three planes of idea plane, verbal plane and notational plane, one of the core ideas in his analytico-synthetic theory of library classification, are analyzed. This article also suggests the demonstration model for this three planes using the ideas from chemistry ad chemical equation. In the last part, it analyzes the canons for library classification of three planes. These normative principles are basically guiding principles for so-called analytico-synthetic or faceted classification. But they can be a n.0, pplied to most of modern classification. But they can be a n.0, pplied to most of modern classification schemes, especially to semi-enumerative schemes including DDC, KDC, etc. so that they can improve the schemes. From this regard, these principles can also be helpful to the KDC, on the verge of the revision of its fourth edition.

• 정보처리학회논문지A
• /
• 제17A권6호
• /
• pp.265-274
• /
• 2010

공정기술 발달로 인해 칩 내부 집적도가 크게 증가하면서 내부 연결망이 멀티코어 프로세서의 성능 향상을 제약하는 주된 원인이 되고 있다. 내부 연결망에서의 지연시간으로 인한 프로세서 성능 저하 문제를 해결하기 위한 방안 중 하나로 3차원 적층 구조 설계 기법이 최신 멀티코어 프로세서를 설계하는데 있어서 큰 주목을 받고 있다. 3차원 적층 구조 멀티코어 프로세서는 코어들이 수직으로 쌓이고 각기 다른 층의 코어들은 TSV(Through-Silicon Via)를 통해 상호 연결되는 구성으로 설계된다. 2차원 구조 멀티코어 프로세서에 비해 3차원 적층 구조 멀티코어 프로세서는 내부 연결망의 길이를 감소시킴으로 인해 성능 향상과 전력소모 감소라는 장점을 가진다. 하지만, 이러한 장점에도 불구하고 3차원 적층 구조 설계 기술은 증가된 전력 밀도로 인해 발생하는 프로세서 내부 온도 상승에 대한 적절한 해결책이 마련되지 않는다면 실제로는 멀티코어 프로세서 설계에 적용되기 어렵다는 한계를 지니고 있다. 본 논문에서는 3차원 멀티코어 프로세서를 설계하는데 있어서 온도 상승 문제를 해결하기 위한 방안 중 하나인 플로어플랜 기법을 다양하게 적용해 보고, 기법 적용에 따른 프로세서의 성능, 전력효율성, 온도에 대한 상세한 분석 결과를 알아보고자 한다. 실험 결과에 따르면, 본 논문에서 제안하는 온도를 고려한 3가지 플로어플랜 기법들은 3차원 멀티코어 프로세서의 온도 상승 문제를 효과적으로 해결함과 동시에, 플로어플랜 변경으로 데이터 패스가 바뀌면서 성능이 저하될 것이라는 당초 예상과는 달리, 온도 하락으로 인해 동적 온도 제어 기법의 적용 시간이 줄어들면서 성능 또한 향상시킬 수 있음을 보여준다. 이와 함께, 온도 하락과 실행 시간 감소로 인해 시스템에서의 전력 소모 또한 줄일 수 있을 것으로 기대된다.

• Nuclear Engineering and Technology
• /
• 제56권6호
• /
• pp.2019-2028
• /
• 2024

Advanced nuclear reactors, especially the newly developed small and micro-reactors have complex neutron spectrum, which makes the deterministic reactor core calculations sensitive to the energy group structure of few-group cross-sections. To avoid significantly increasing the cost of energy discretization in the core calculation, two energy group structures with 31 groups and 33 groups were adopted for typical thermal and fast reactor cores, respectively. Then, an adaptive scheme of group division for reactor cores with a medium neutron spectrum was proposed. The works were based on the full spectrum nuclear reactor analysis code SARAX/TULIP. An equivalent one-dimensional model of the core was proposed to capture the key neutron spectrum features of the reactor core. Such features were used to adaptively determine a few-group structure for the following reactor core calculations. Then, the neutron spectrum in different zones with more details was calculated. With this spectrum, the cross-sections were condensed into the determined energy groups. Three tests based on different neutron spectrum were calculated to verify the schemes. The results show that using the adaptive energy group division scheme, the following core calculation can meet the accuracy requirement of different reactors with different neutron spectra.

• 정보처리학회논문지C
• /
• 제13C권5호
• /
• pp.659-668
• /
• 2006

차세대 네트워크에서 이동성 지원을 위해 다양한 방식이 제시되고 있으나 이들의 성능이 상대적으로 비교된 바는 없다. 따라서, 본 논문에서는 MIPv6, IP2, QMMF, IMS 방식의 위치 등록 과정과 핸드오버 절차를 분석하고, 성능 분석을 위한 파라미터 및 시뮬레이션 방법을 이용하여 각각의 절차에 대한 등록 및 핸드오버 지연 시간과 비용 측면에서 성능을 비교 분석하였다. 결과적으로 위치관리자(Location Manager)가 코어 네트워크 상의 중심에 위치한 IP2가 다른 메커니즘들에 비해 가장 좋은 성능을 보임을 확인하였다.

• 한국콘크리트학회:학술대회논문집
• /
• 한국콘크리트학회 1995년도 봄 학술발표회 논문집
• /
• pp.201-206
• /
• 1995

This paper presents the experimental study on crack control of core-wall placed with 700kg/$\textrm{cm}^2$ Ultra high strength concrete. The thermal sensors were installed into the core-walls prior to the concrete casting to measure the heat of hydration and atmospheric temperature whose difference might cause the initial crack. Several curing schemes were taken for each basement floor 8 thru 6 to examine the influence of curing method on the crack width, total crack length and the number of crack occurred. This paper demonstrates that the proper curing scheme have a great favorable effect on the initial crack control on the structural elements with noticiable reduction in crack width.