DOI QR코드

DOI QR Code

GP-GPU based Parallelization for Urban Terrain Atmospheric Model CFD_NIMR

도시기상모델 CFD_NIMR의 GP-GPU 실행을 위한 병렬 프로그램의 구현

  • Kim, Youngtae (Department of Computer Science & Engineering, Gangneung-Wonju National University) ;
  • Park, Hyeja (Forecast Research Division, National Institute of Meteological Research) ;
  • Choi, Young-Jeen (Weather Information Service Engine Foundation)
  • Received : 2014.01.11
  • Accepted : 2014.03.10
  • Published : 2014.04.30

Abstract

In this paper, we implemented a CUDA Fortran parallel program to run the CFD_NIMR model on GP-GPU's, which simulates air diffusion on urban terrains. A GP-GPU is graphic processing unit in the form of a PCI card, and a general calculation accelerator to perform a large amount of high speed calculations with low cost and electric power. The GP-GPU gives performance enhancement of speed by 15 times to compare the Nvidia Tesla C1060 GPU with Intel XEON 2.0 GHz CPU. In addition, the program on a GP-GPU shows efficient performance compared to an MPI parallel program on multiple CPU's. It is expected that a proposed programming method on the GP-GPU parallel program can be used for numerical models with a similar structure.

본 논문은 도시기상모델인 전산유체역학모델(CFD_NIMR)을 GP-GPU에서 실행시키기 위해 CUDA Fortran 병렬프로그램을 구현하였다. GP-GPU는 원래 PCI 카드 형태의 그래픽 처리 장치이지만 저비용, 저전력으로 대량의 계산을 초고속으로 수행할 수 있는 일반 계산 가속기이다. 모델을 단일 Intel XEON 2.0 GHz CPU에서 실행한 결과와 Nvidia Tesla C1060 GPU에서 실행한 성능을 비교하였을 때 GP-GPU에서 15배 정도의 빠른 속도를 보였다. 또한 다중 CPU를 사용한 MPI 병렬프로그램과 비교한 경우에도 GP-GPU에서 보다 더 효율적인 성능을 보였다. 본 논문에서 제시한 프로그램 방식은 유사한 구조를 가진 수치모델을 GP-GPU 병렬 프로그램으로 구현하는데 쉽게 적용할 수 있을 것으로 기대한다.

Keywords

References

  1. Meuer, H., The future of HPC, Scientific Computing World: June/July 2009.
  2. The Top Trends in High Performance Computing, The Top 500 Report, Top 500 Supercomputer Sites, 23 June 2009.
  3. NVIDIA CUDA Compute United Device Architecture Reference Manual Version 5.0, NVIDIA Corporation, 2012.
  4. Nvidia, CUDA Programming Guide 4.2.
  5. CUDA Fortran Programming Guide and Reference, The Portland Group, 2012.
  6. National Institute of Meteorological Research, Diagnosis of characteristics of local meteorology and development of techniques for the meteorological environmental impact assessment (I), 218 pp, 2006.
  7. Budruk, R., D. Anderson, T. Shanley, and J. Winkles, PCI Express System Architecture, Mind share PC system architecture, Addison-Wesley, pp 1120, 2003.
  8. Amdahl, G., Validity of the Single Processor Approach to Achieving Large-Scale Computing Capabilities", AFIPS Conference Proceedings (30): 483-485, 1967. (2013)
  9. Youngtea Kim, Yong Hee Lee, and Kwan-Young Chung, WRF Physics Models Using GP-GPUs with CUDA Fortran, Atmosphere. Korean Meteorological Society, Vol. 23, No. 2, 231-235, 2013.
  10. Molnar Jr., T. Szakaly, R. Meszaros, and I. Lagzi, Air pollution modeling using a graphics processing unit with CUDA, Comput. Geosci. 36(5), 105-112., 2010.
  11. Min-Wook Kim,Young-Jean Choi, and Youngtae Kim, Hybrid Parallelization for High Performance of CFD_NIMR Model, Atmosphere. Korean Meteorological Society, Vol. 22, No. 1, 109-115, 2012.