期刊文献+

CPU-GPU混合平台上动态场景光线跟踪的研究 被引量:5

Investigation of ray tracing in dynamic scenes on CPU-GPU hybrid paltform
在线阅读 下载PDF
导出
摘要 提出了一种动态场景光线跟踪新方法,能有效地调度CPU和GPU的运行,提高渲染速度。根据加速结构kd-tree的特点,将其分成上层部分和下层部分,上层部分由于并行性较小,由CPU创建;而下层部分并行性较大,由GPU创建,提高动态场景加速结构的创建速度。同时充分利用CPU和GPU两个运算平台的特点,有效调度两者的运行,隐藏部分运算时间,进一步提高动态场景的渲染速度。实验结果表明,在安装了GeForce285GTX的PC机上,高真实感地交互渲染了包含11k三角面片的Kitchen动态场景。 A new method of ray tracing in dynamic scenes is proposed.It can effectively schedule the operation of CPU and GPU to enhance rendering speed.According to the characteristics of kd-tree,it is divided into the upper part and lower part.Because the parallelism of upper part is small,it is created by CPU,and the lower part is created by GPU.The speed to create kd-tree in dynamic scence is improved.Taking advantage of both CPU and GPU,the operation of the two computing platform is scheduled effectively.Part of the computing time is hidden to enhance the rendering speed.The results show on the PC with GeForce285 GTX,the Kitchen scence with 11k triangles can be rendered rapidly.
出处 《计算机工程与应用》 CSCD 2012年第21期151-154,159,共5页 Computer Engineering and Applications
基金 江苏省自然科学基金(No.BK2009351)
关键词 光线跟踪 KD-TREE 图形处理器(GPU) CPU ray tracing kd-tree Graphics Processing Uni(tGPU) CPU
  • 相关文献

参考文献9

  • 1Glassner A S.An introduction to ray tracing[M].London, UK:Academic Press Ltd, 1989.
  • 2Wald I,Mark W R,Gunther J, et al.State of the art in ray tracing animated scenes[J].Computers & Graphics, 2008,32(1):3-13.
  • 3Choi B,Komuravelli R, Lu V, et al.Parallel SAH k-D Tree construction for fast dynamic scene ray tracing[R]. 2009.
  • 4Shevtsov M, Soupikov A, Kapustin A.Highly parallel fastkd-tree construction for interactive ray tracing of dynamic scenes[J].Computer Graphics Forum, 2007,26 ( 3 ) : 395-404.
  • 5Budge B,Anderson J,Garth C,et al.A straightforward CUDA implementation for interactive ray-tracing[C]//IEEE Symposium on Interactive Ray Tracing, Los Angeles, CA, 2008: 178-185.
  • 6Wald I, Benthin C.Interactive rendering with coherent ray tracing[J].Computer Graphics Forum, 2001,20 ( 3 ) : 153 - 164.
  • 7Thrane N, Simonsen L O.A comparison of acceleration structures for GPU assisted ray tracing[D].University of Aarhus. 2005.
  • 8Hunt W,Mark W R, Stoll G.Fast kd-tree construction with an adaptive error-bounded heuristic[C]//IEEE Sympo- sium on Interactive Ray Tracing, Salt Lake City,UT, 2006 : 81-88.
  • 9BART: a Benchmark for Animated Ray Tracing[EB/OL] http://www.ce.chalmers.se/research/group/graphics/BART/.

同被引文献43

  • 1郑长伟,刘高攀,郭齐胜.基于OpenGVS4.3的光照效果研究[J].计算机仿真,2005,22(7):198-200. 被引量:3
  • 2韩慧健,徐琳.基于图像的三维光照效果的动态重现[J].计算机应用,2005,25(9):2123-2125. 被引量:1
  • 3Volkov V,Demmel J W. Benchmarking gpus to tune dense linear al- gebra. Proceedings of the 2008 ACM/IEEE Conference on Supercom- puting, 2008 : 1 -11.
  • 4Watkins D S. Fundamentals of matrix computations Wiley and Sons, 2013.
  • 5New York : John Baltimore : Johns Karlsson L, K:gstr6m B. Parallel two-stage reduction to Hessenberg form using dynamic scheduling on shared-memory architectures. Par- allel Computing, 2011 ;37(12): 771-782.
  • 6Tomov S, Nath R, Dongarra J. Accelerating the reduction to upper Hessenberg, tridiagonal, and bidiagonal forms through hybrid GPU- based computing. Parallel Computing, 2010;36( 12): 645-654.
  • 7Muramatsu J, Zhang S, Yamamoto Y. Acceleration of hessenberg re- duction for nonsymmetric eigenvalue problems using GPU. First Inter- national Conference on Networking and Computing, 2010 : 125 -129.
  • 8Golub G H, Van Loan C F. Matrix computations. Hopkins University Press, 2013.
  • 9Chandrasekar J, Kim I S, Bernstein D S, et al. Reduced-Rank Unscent- ed Kalman filtering using Cholesky-based decomposition [ C ]//Ameri- can Control Conference, June,2008 : 1274 - 1279.
  • 10Yu H, Chung C Y, Wong K P, et al. Probabilistic Load Flow Evaluation With Hybrid Latin Hypereube Sampling and Cholesky Decomposition [ J]. IEEE Transactions on Power Systems,2009,24 (2) :661 - 667.

引证文献5

二级引证文献12

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部