• 전자공학회논문지A
• /
• 제31A권3호
• /
• pp.53-67
• /
• 1994

This paper describes the design of a high-level synthesis system. SODAS-VP. which automatically generates hardwares executing operation sequences in pipelined fashion.Target architecture and clocking schemes to drive pipelined datapath are determined, and the handling of pipeline hazards which degrade the performance of pipeline is considered. Partitioning of an operation into load, operation, and store stages, each of which is executed in partitiones control step, is performend. Pipelinecl hardware is generated by handling pipeline hazards with internal forwarding or delay insertion techniques in partitioning process and resolving resource conflicts among the partitioned control steps with similarity measure as a priority function in module allocation process. Experimental results show that SODAS-VP generates hardwares that execute faster than those generated by HAL and ALPS systems. SODAS-VP brings improvement in execution speed by 17.1% and 7.4% comparing with HAL and ALPS systems for a MCNC benchmark program, 5th order elliptical wave filter,respectively.

• 한국전산유체공학회지
• /
• 제13권3호
• /
• pp.69-78
• /
• 2008

The high-order numerical method based on the adaptive mesh refinement(AMR) on the quadrilateral unstructured grids has been developed in this paper. This adaptive-grid method, originally developed with MUSCL-TVD scheme, is now extended to the WENO (weighted essentially no-oscillatory) scheme with the Runge-Kutta time integration of fifth order in spatial and temporal accuracy. The multidimensional interpolation was studied in the preliminary research, which allows us to maintain the same order of accuracy for the computation of numerical flux between two adjacent cells of different levels. Some standard benchmark tests are done to validate this method for checking the overall capacity and efficiency of the present adaptive-grid technique.

• Structural Engineering and Mechanics
• /
• 제36권2호
• /
• pp.225-241
• /
• 2010

In this study, a hybrid search algorithm combining genetic programming with orthogonal least squares (GP/OLS) is utilized to generate prediction models for compressive strength of high performance concrete (HPC) mixes. The GP/OLS models are developed based on a comprehensive database containing 1133 experimental test results obtained from previously published papers. A multiple least squares regression (LSR) analysis is performed to benchmark the GP/OLS models. A subsequent parametric study is carried out to verify the validity of the models. The results indicate that the proposed models are effectively capable of evaluating the compressive strength of HPC mixes. The derived formulas are very simple, straightforward and provide an analysis tool accessible to practicing engineers.

• Journal of Computing Science and Engineering
• /
• 제11권4호
• /
• pp.130-141
• /
• 2017

Feature selection is one of the most challenging problems of pattern recognition and data mining. In this paper, a feature selection algorithm based on an improved version of binary differential evolution is proposed. The method simultaneously optimizes two feature selection criteria, namely, set approximation accuracy of rough set theory and relational algebra based derived score, in order to select the most relevant feature subset from an entire feature set. Superiority of the proposed method over other state-of-the-art methods is confirmed by experimental results, which is conducted over seven publicly available benchmark datasets of different characteristics such as a low number of objects with a high number of features, and a high number of objects with a low number of features.

• 한국통신학회논문지
• /
• 제17권4호
• /
• pp.424-432
• /
• 1992

본 논문에서는 고위영역합성(High-level synthesis)의 가장 중요한 과제인 스케듈링 문제를 다루었다. 스케듈링 문제에 대한 접근 방식으로서, IP(integer programming)을 이용한 방식을 택하였다. 본 논문에서는 특히, 가용(可用)자원이 제한된 상태에서의 스케듈링이 효율적으로 수행될 수 있는 방법을 연구하여, 임의의 스케듈링방식에 의해 구해진 스케듈링 결과로부터, 주어진 조건을 만족하는 스케듈링을 행할수 있는 방법을 연구하였다. 멀티싸이클 연산의 연산자할당 및 구조적 파이프라이닝을 위한 스케듈링에 중점을 주어, 가능한한 최대의 성능과 최대의 자원공유가 이루어 지도록, 연산자의 특성을 세밀히 분석하였다.

• Journal of the Korean Data and Information Science Society
• /
• 제18권1호
• /
• pp.161-174
• /
• 2007

In this paper, we propose a high performance database cluster system called HyperDB to process OLAP queries efficiently. HyperDB is a virtual database system running on top of internet-connected PCs; the PCs are used for their own purpose at ordinary times, but they are able to participate in the database cluster system at non-office hours. We propose fully logical replication technique and optimal parallel intra-query routing technique for extensibility and performance. Experiment for TPC-R benchmark shows significant performance upgrade compared with conventional approaches.

• 한국전산유체공학회:학술대회논문집
• /
• 한국전산유체공학회 2006년도 추계 학술대회논문집
• /
• pp.47-50
• /
• 2006

In the aeroacoustic application of computational fluid dynamics, the physical phenomena like the crackle in the unsteady compressible jets should be based on very time-accurate numerical solution. The accuracy of the present numerical scheme is extended to the fifth order, using the WENO filter to the sixth-order central difference computation. However, the computational capacity is very restricted by the environment of computational power, so therefore the quadrilateral adaptive grids technique is introduced for this high-order accuracy scheme. The first problem is the multi-dimensional interpolation between fine and coarse grids. Some general benchmark problems are solved to show the effectiveness of this method.

• 소성∙가공
• /
• 제21권6호
• /
• pp.384-388
• /
• 2012

Sheet metal forming processes induce residual stress in the final product due to plastic deformation. The residual stress leads to elastic recovery of the formed part called springback, which causes shape errors in the final product. This error is a serious issue, especially for high strength steels, which are widely used in auto-body structures. Therefore, the evaluation of the amount of springback becomes critical for high strength steels. This paper investigates the springback behavior of DP780 steel sheets after the U-bending process using the geometry of the standard U-shape tool from the NUMISHEET'93 benchmark problem. The amounts of springback were measured as a function of the intrusion direction, forming speed and blank holding force.

• Computers and Concrete
• /
• 제1권2호
• /
• pp.189-209
• /
• 2004

Several extensions to the Thelandersson phenomenological model for concrete under transient high temperatures are explored. These include novel expressions for the temperature degradation of the elastic modulus and the temperature dependency of the coefficient of the free thermal strain. Furthermore, a coefficient of thermo mechanical strain is proposed as a bi-linear function of temperature. Good qualitative agreement with various test results taken from the literature is demonstrated. Further extensions include the effects of plastic straining and temperature dependent Poisson's ratio. The models performance is illustrated on several simple benchmark problems under uniaxial and biaxial stress states.

• 전자공학회논문지A
• /
• 제31A권6호
• /
• pp.177-187
• /
• 1994

In the design of modern computer systems, the speed gap between the CPUs and DRAMs has been a major concern. To relieve this problem at a low cost, several new DRAM architectures have been proposed. This study is aimed at evaluating quantitatively the impact of the new DRAM architectures (synchronous DRAM. dual-RAS synchronous DRAM, and enhanced DRAM) on the memory system performance. We developed a cache and memory simulator and performed various experiments using the traces generated from four benchmark programs. The simulation results show that the new DRAM architectures offer a better performance than a conventional one by 5~30% in a low cost system and their improvement in a high performance system is less than 1%. However, for resonable multiprogramming workoads, additional performance improvement of about 10~28% is expected in a high performance system while 1~3% in a low cost system.