• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 1998년도 추계종합학술대회 논문집
• /
• pp.463-466
• /
• 1998

본 논문에서는 메시지 패싱 전송을 기반으로 하는 병렬 처리 시스템의 상호연결망 구조와 특성을 조사하고 비교한다. 특히, 상용 시장에서 널리 알려진 대표적인 병렬 처리 시스템의 상호연결망 특성과 ETRI에서 개발된 고속 병렬 컴퓨터(SPAX)의 계층 크로스바 상호 연결망(Xcent-Net) 특성을 상호 비교한다. 메시지 패싱 전송 기반의 상호연결망은 일반적으로 확장성이 우수하여 대규모 병렬 처리 시스템을 구축하는데 유리하다. Cray T3E 시스템, Intel ASCI TFLOPS 시스템, Tandem Himalaya S70000 시스템, IBM RS6000 SP2 시스템등은 메시지 패싱 상호연결망을 기반으로 수백개에서 수천개의 대규모 프로세서를 연결한 병렬 처리 시스템이다. ETRI SPAX 시스템은 Xcent-Net 메시지 패싱 상호연결망을 기반으로 최대 256개 프로세서를 연결한 고속 병렬 처리 시스템으로 우수한 확장성과 높은 성능을 제공한다. 본 논문에서는 상호연결망의 구조와 함께 라우팅 스위치 구조 및 특성을 중심으로 전송 지연시간, 그리고 노드당 전송 대역폭 특성을 비교한다.

• 한국전산유체공학회:학술대회논문집
• /
• 한국전산유체공학회 2000년도 추계 학술대회논문집
• /
• pp.155-162
• /
• 2000

In this paper, preconditioning method is parallelized on fast-ethernet PC cluster. The algorithm is based on scaling the pressure terms in the momemtum equations and preconditioning the conservation equations to circumvent numerical difficulties at low Mach numbers. Parallelization is performed using a domain decomposition technique(DDT) and message passing between sub-domains are taken from the MPI library. The results are shown to have good convergence properties at all Mach number on the circular arc Bump and are capable of reasonable predicting two-dimensional turbulent flows on DCA compressor cascade.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1991년도 한국자동제어학술회의논문집(국제학술편); KOEX, Seoul; 22-24 Oct. 1991
• /
• pp.1633-1638
• /
• 1991

A CAI system is developed to support the Instruction of an aircraft conceptual design for aeronautical engineering students. Three system concepts are proposed and an Object-Oriented approach is applied to construct the system. The system has three major functions to perform a conceptual design: (1) the system stores modular data and empirical formulas used for a wide range of aircraft design tasks from light aircraft to long range airliners. (2) Implementation of modules by message passing makes it easy to realize the various design tasks required for different design requirements. (3) The system allows users to study trade-off among the requirements. The system has a graphical user Interface which allows users to communicate with the system interactively. The effectiveness of the system Is demonstrated through some case studies.

• 한국유체기계학회 논문집
• /
• 제1권1호
• /
• pp.72-80
• /
• 1998

Computational analysis of incompressible flows by numerically solving Navier-Stokes equations using multi-block finite volume method is conducted on a parallel computing system. Numerical algorithms adopted in this study $include^{(1)}$ QUICK upwinding scheme for convective $terms,^{(2)}$ central differencing for other terms $and^{(3)}$ the second-order Euler differencing for time-marching procedure. Structured grids are used on the body-fitted coordinate with multi-block concept which uses overlaid grids on the block-interfacing boundaries. Computational code is parallelized on the MPI environment. Numerical accuracy of the computational method is verified by solving a benchmark test case of the flow inside two-dimensional rectangular cavity. Computation in the axial compressor cascade is conducted by using 4 PE's md, as results, no numerical instabilities are observed and it is expected that the present computational method can be applied to the turbomachinery flow problems without major difficulties.

• Journal of electromagnetic engineering and science
• /
• 제10권3호
• /
• pp.166-170
• /
• 2010

To numerically calculate electromagnetic scattering from the electrical-large three-dimensional(3D) objects, the high-frequency approaches have been usually applied, but the accuracy and feasibility of these geometrical and physical optics(GO-PO) approaches, to some extent, are remained to be improved. In this paper, a new framework is developed for calculation of the near-field scattering field of an electrical-large 3D target by using a multilevel fast multipole algorithm(MLFMA) and generation of radar images by using a fast back-projection(FBP) algorithm. The MPI(Message Passing Interface) parallel computing is carried out to multiply the calculation efficiency greatly. Finally, a simple example of perfectly electrical conducting(PEC) patch and a canonical case of Fighting Falcon F-16 are presented.

• 센서학회지
• /
• 제27권3호
• /
• pp.143-149
• /
• 2018

In this paper, we introduce an edge-spike modulation method for visible light communication (VLC). This method is effective in preventing LED flicker and 120 Hz noise interference in base-band VLC. In the VLC transmitter, edge-spikes are generated by passing the digital data through a simple RC-high pass filter (HPF). The LED modulation of the edge-spikes does not change the average power of the LED light; thus it prevents LED flicker. In the VLC receiver, the 120 Hz noise from other lighting lamps is easily eliminated by RC-HPF, while the edge-spike signal is detected normally. In our experiment, the message of an air-quality sensor was successfully transmitted using edge-spike modulation. This structure is useful in constructing, e.g., wireless gas monitoring sensor systems to warn and prevent harmful gas leakage accidents in buildings using LED light.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송공학회 2014년도 하계학술대회
• /
• pp.77-78
• /
• 2014

A new MCMC method for optimization is presented in this paper, which is called the scanline block Gibbs sampler. Due to its slow convergence speed, traditional Markov chain Monte Carlo (MCMC) is not widely used. In contrast to the conventional MCMC method, it is more convenient to parallelize the scanline block Gibbs sampler. Since The main part of the scanline block Gibbs sampler is to calculate message between each edge, in order to accelerate the calculation of messages passing in scanline sampler, it is parallelized in GPU. It is proved that the implementation on GPU is faster than on CPU based on the experiments on the OpenGM2 benchmark.

• 한국산업정보학회논문지
• /
• 제4권4호
• /
• pp.88-93
• /
• 1999

선형독립인 소수의 벡터들을 직교화 시키는 방법인 Cram-Schmidt 알고리즘, 몇 개의 극단적인 고유치를 구하는 방법인 란초스 알고리즘방법 등이, 한번의 반복동안 한번의 동기점만 가지도록 재구성된다. 즉 메시지 패싱 분산시스템에서 모든 프로세서들 사이에 한번의 통신만 요구되도록 알고리즘을 변화시킨다. 이러한 알고리즘들은 재구성되었다고 하며, 전통적인 방법에 비하여 더 나은 병렬성질을 가지게 된다.

• 한국전산유체공학회:학술대회논문집
• /
• 한국전산유체공학회 2004년도 춘계 학술대회논문집
• /
• pp.172-177
• /
• 2004

In the present work, we have interests on the modification of parallel implemented with MPI(Message Passing Interface) programming method, 3-Dimensional, unsteady, incompressible Navier-Stokes equation solver to analyze the low-Reynolds number flow In order to accurate calculation aerodynamic coefficients in low-Reynolds number flow field, we modified the two-equation turbulence model. This paper describes the development and validation of a new two-equation model for the prediction of flow transition. It is based on Mentor's low Reynolds $\kappa-\omega$ model with modifications to include Total Stresses Limitation (TSL) and Separation Transition Trigger (STT)

• 한국정보기술응용학회:학술대회논문집
• /
• 한국정보기술응용학회 2005년도 6th 2005 International Conference on Computers, Communications and System
• /
• pp.13-16
• /
• 2005

It would be desirable to have methods for specific problems, which have low communication costs compared to the computation costs, and in specific applications, algorithms need to be developed and mapped onto parallel computer architectures. Main memory access for shared memory system or global communication in message passing system deteriorate the computation speed. In this paper, it is found that the m-step generalization of the block Lanczos method enhances parallel properties by forming m simultaneous search direction vector blocks. QR factorization, which lowers the speed on parallel computers, is not necessary in the m-step block Lanczos method. The m-step method has the minimized synchronization points, which resulted in the minimized global communications compared to the standard methods.