Adaptive Load Balancing Scheme using a Combination of Hierarchical Data Structures and 3D Clustering for Parallel Volume Rendering on GPU Clusters

계층 자료구조의 결합과 3차원 클러스터링을 이용하여 적응적으로 부하 균형된 GPU-클러스터 기반 병렬 볼륨 렌더링

  • 이원종 (연세대학교 컴퓨터과학과) ;
  • 박우찬 (세종대학교 인터넷공학과) ;
  • 한탁돈 (연세대학교 컴퓨터과학과)
  • Published : 2006.02.01

Abstract

Sort-last parallel rendering using a cluster of GPUs has been widely used as an efficient method for visualizing large- scale volume datasets. The performance of this method is constrained by load balancing when data parallelism is included. In previous works static partitioning could lead to self-balance when only task level parallelism is included. In this paper, we present a load balancing scheme that adapts to the characteristic of volume dataset when data parallelism is also employed. We effectively combine the hierarchical data structures (octree and BSP tree) in order to skip empty regions and distribute workload to corresponding rendering nodes. Moreover, we also exploit a 3D clustering method to determine visibility order and save the AGP bandwidths on each rendering node. Experimental results show that our scheme can achieve significant performance gains compared with traditional static load distribution schemes.

대용량 볼륨 데이타를 가시화하는 효과적인 방법인 후-정열 병렬 렌더링은 부하균형에 의해 성능이 결정된다. 기존의 정적 데이타 분할 방법은 태스크 병렬성만의 관점에서는 자기균형을 쉽게 얻을 수 있었지만, 데이타 내부의 빈 공간을 고려하지 않았기 때문에 데이타 병렬성의 관점에서는 심각한 불균형을 초래할 수 있었다. 본 논문은 태스크 병렬성과 데이타 병렬성이 함께 고려된, 적응적이며 확장적인 부하 균형 기법을 제안한다. 우리는 계층적 자료 구조인 옥트리와 BSP-트리를 효과적으로 결합하여 볼륨 데이타의 실제 영역만을 추출하여 렌더링 노드들로 균등하게 분산시켰으며, 각 렌더링 노드들에서는 3차원 클러스터링 알고리즘을 적용하여 렌더링 순서를 효과적으로 결정하였다. 제안하는 방법은 기존의 정적 데이타 분산 기법에 비해 최대 22배의 병렬성을 높였고 동기화 비용을 낮추어 렌더링 성능을 크게 향상시켰음을 실험을 통해 알 수 있었다.

Keywords

References

  1. S. Molnar, M. Cox, D. Ellsworth, and H. Fuchs., 'A sorting classification of parallel rendering,' IEEE Computer Graphics and Applications, Vol. 14, No.4, pp. 23-32, 1994 https://doi.org/10.1109/38.291528
  2. R.J, Karia., 'Load balancing of parallel volume rendering with scattered decomposition,' In Proc. Scalable High Performance Computing Conference '94, pp. 252-258, 1994 https://doi.org/10.1109/SHPCC.1994.296651
  3. K.L. Ma, J, Painter, C. Hansen, and M. Krogh., 'A data distributed parallel rendering algorithm for ray-traced volume rendering,' In Proc. Parallel Rendering Symposium '93, pp. 117-126, 1993 https://doi.org/10.1145/166181.166183
  4. K.L. Ma and S. Parker., 'Massively parallel software rendering for visualizing large-scale data sets,' IEEE Computer Graphics and Applications, Vol. 21, No.4, pp. 72-83, 2001 https://doi.org/10.1109/38.933526
  5. C. Silva, A. Kaufman, and C. Pavlakos., 'PVR: High-performance volume rendering,' In Proc. IEEE Computational Science and Engineering '96, pp. 18-28, 1996 https://doi.org/10.1109/99.556509
  6. C. Wang, J. Gao, and H.W Shen., 'Parallel multiresolution volume rendering of large data sets with error-guided load balancing,' In Proc. Symposium on Parallel Graphics and Visualization '04, pp. 18-25, 2004
  7. J.M. Kniss, P. McCormick, A. McPherson, J. Ahrens, J. Painter, A. Keaheyand, and C. Hansen., 'Tv-Rex: Interactive texture-based volume rendering for large data sets,' IEEE Computer Graphics and Applications, Vol. 21, No.4, pp. 52-61, 2001 https://doi.org/10.1109/38.933524
  8. M. Magallon, M. Hopf, and T. Ertl., 'Parallel volume rendering using PC graphics hardware,' In Proc. Pacific Graphics '01, pp. 384-389, 2001 https://doi.org/10.1109/PCCGA.2001.962895
  9. S. Muraki, E.B. Lum, KL. Ma, M. Ogata, and X. Liu., 'A PC cluster system for simultaneous interactive volumetric modeling and visualization,' In Proc. Parallel Visualization and Graphics '03, pp, 95-102, 2003 https://doi.org/10.1109/PVGS.2003.1249047
  10. M. Strengert, M. Magallon, D.Weiskopf, S. Guthe, and T. Ertl., 'Hierarchical visualization and compression of large volume datasets using GPU clusters,' In Proc. Symposium on Parallel Graphics and Visualization '04, pp. 1-7, 2004
  11. D. Bartz, B. Schneider, and C. Silva., 'Rendering and visualization in parallel environments,' In Proc. SIGGRAPH 2000 Coursenote, 2000
  12. K.L. Ma, J.S. Painter, C. Hansen, and M.F. Krogh., 'Parallel volume rendering using bnary-swaji image composition,' IEEE Computer Graphics and Applications, Vol. 14, No.4, pp.59-68, 1994 https://doi.org/10.1109/38.291532
  13. J. Nonaka, N. Kukimoto, N. Sakamoto, H. Hazama, Y. Watashiba, X. Liu, M. Ogata, M. Kanazawa, and K. Koyamada., 'Hybrid hardware accelerated image composition for sort-last parallel rendering on graphics clusters with commodity image compositor,' In Proc. Symposium on Volume Visualization and Graphics '04, pp. 17-24, 2004 https://doi.org/10.1109/VV.2004.4
  14. M. Ogata, S. Muraki, X. Liu, and KL. Ma., 'The design and evaluation of a pipelined image compositing device for massively parallel volume rendering,' In Proc. Workshop on Volume Graphics '03, pp. 61 -68, 2003 https://doi.org/10.1145/827051.827060
  15. J. Kruger and R. Westermann., 'Acceleration techniques for GPU based volume rendering,' In Proc. IEEE Visualization '03, pp. 287-292, 2003. https://doi.org/10.1109/VIS.2003.10001
  16. W. Li, K Mueller, and A. Kaufman., 'Empty space skipping and occlusion clipping for texture- ?based volume rendering,' In Proc. IEEE Visualization '03, pp. 317-324, 2003 https://doi.org/10.1109/VISUAL.2003.1250388
  17. M.J. Berger and I. Rigoutsos., 'An algorithm for point clustering and grid generation,' IEEE Transaction on Systems, Man, and Cybernetics, Vol. 21, No.5, 1991 https://doi.org/10.1109/21.120081
  18. T.W. Crockett., 'Parallel rendering,' Encyclopedia of Computer Science and Technology, Vol. 34, No. 19, pp. 335-371, 1996
  19. J. Gao, J. Huang, H.W. Shen, and A. Kohl., 'Visibility culling using plenoptic opacity functions for large data visualization,' In Proc. IEEE Visualization '03, pp. 341-348, 2003 https://doi.org/10.1109/VISUAL.2003.1250391
  20. D. Bartz, W. StraBer, R. Grosso, and T. Ertl., 'Parallel construction and isosurface extraction of recursive tree tructures,' In Proc. WSCG, '98, 1998
  21. J. Bielak E.J. Kim and O. Ghattas., 'Large-scale northridge earthquake simulation using octreebased multiresolution mesh method,' In Proc. ASCE 16th Engineering Mechanics Conference '03, 2003
  22. L.A. Freitag and R.M. Loy., 'Adaptive, multiresolution visualization of large data sets using parallel octrees,' In Proc. ACM/IEEE conference on Supercomputing '99, 1999
  23. W.G. Aref and H. Samet., 'An algorithm for perspective viewing of objects represented by octrees,' Computer Graphics Forum, Vol. 14, No. 1, pp. 59-66, 1995 https://doi.org/10.1111/1467-8659.1410059
  24. R. Kahler, M. Simon, and H.C. Hege., 'Interactive volume rendering of large sparse data sets using adaptive mesh refinement hierarchies,' IEEE Transaction on Visualization and Computer Graphics, Vol. 9, No.3, pp. 341-351, 2003 https://doi.org/10.1109/TVCG.2003.1207442