• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.42 no.12
• /
• pp.61-70
• /
• 2005

In this paper, we propose an efficient global placement algorithm which is an enhanced version of Hybrid Placer$^{[25]}$, a standard cell placement tool, which uses a middle-down approach. Combining techniques used in the well-known partitioner hMETIS and the RBLS(Relaxation Based Local Search) in Hybrid Placer improves the quality of global placements. Partitioning techniques of hMETIS is applied in a top-down manner and RBLS is used in each level of the top-down hierarchy to improve the global placement. The proposed new approach resolves the problem that Hybrid Placer seriously depends on initial placements and it speeds up without deteriorating the placement quality. Experimental results prove that solutions generated by the proposed method on the MCNC benchmarks are comparable to those by FengShui which is a well known placement tool. Compared to the results of the original Hybrid Placer, new method is 5 times faster on average and shows improvement on bigger circuits.

• Proceedings of the Korean Information Science Society Conference
• /
• 2004.04a
• /
• pp.967-969
• /
• 2004

본 논문에서는 표준 셀 배치기 Mongrel을 소개하고 두 가지 분할 기법(FM, hMETIS)을 통해 효율적인 초기 배치를 얻음으로써 Mongrel의 성능을 개선한다. Mongrel은 middle-down 방법론을 채택한 전역 배치와 상세 배치의 2단계 표준 셀 배치기이며 전역 배치 단계에서는 RBLS(Relaxation-Based Local Search)를, 그리고 상세 배치 단계에서는 최적 인터리빙(optimal interleaving) 기법을 통해 최적화 알고리즘을 수행한다. MCNC 벤치마크 회로를 이용한 실험을 통해 Mongrel의 성능을 평가하고 효율적인 초기배치가 최종 결과에 미치는 영향을 분석한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2004.10a
• /
• pp.742-744
• /
• 2004

본 논문에서는 표준 셀 배치기 Mongrel의 성능을 개선하기 위해 사용된 다양한 기법에 관해 살펴보고 top-down방식의 계층적 분할 기법을 이용한 광역 배치(Hierarchical Global Placement)를 제안한다. 계층적 분할 기법을 이용한 광역 배치는 RBLS(Relaxation Based Local Search) 기법과 더불어 Mongrel의 성능 개선에 결정적인 역할을 하고 있으며 분할 기법으로 hMETIS(클러스터링을 이용한 다단계 분할 기법)를 사용한다. 우리는 표준 벤치마크 회로를 이용한 실험을 통해 계층적 분할 기법을 이용한 광역 배치 기법이 안정적이면서 효율적인 배치 결과를 가져옴을 보인다.

• 한국전산유체공학회:학술대회논문집
• /
• 2003.10a
• /
• pp.84-85
• /
• 2003

A parallelized FEM code based on domain decomposition method has been recently developed for a large scale computational fluid dynamics. A 4-step splitting finite element algorithm is adopted for unsteady computation of the incompressible Navier-Stokes equation, and Smagorinsky LES(Large Eddy Simulation) model is chosen for turbulent flow computation. Both METIS and MPI library are used for domain partitioning and data communication between processors respectively. Tiburon of Hyundai-motor is chosen as the computational model at $Re=7.5{\times}10^{5}$, which is based on the car height. It is confirmed that the drag under road condition is smaller than that of wind tunnel condition.

• 한국전산유체공학회:학술대회논문집
• /
• 2002.05a
• /
• pp.8-15
• /
• 2002

For the large scale computation of turbulent flows around an arbitrarily shaped body, a parallel LES (large eddy simulation) code has been recently developed in which domain decomposition method is adopted. METIS and MPI (message Passing interface) libraries are used for domain partitioning and data communication between processors, respectively. For unsteady computation of the incompressible Wavier-Stokes equation, 4-step splitting finite element algorithm [1] is adopted and Smagorinsky or dynamic LES model can be chosen fur the modeling of small eddies in turbulent flows. For the validation and performance-estimation of the parallel code, a three-dimensional laminar flow generated by natural convection inside a cube has been solved. Then, we have solved the turbulent flow around MIRA (Motor Industry Research Association) model at $Re = 2.6\times10^6$, which is based on the model height and inlet free stream velocity, using 32 processors on IBM SMP cluster and compared with the existing experiment.