Reconfigurable RTGPU's Architecture and Simulation

doi:10.16182/j.issn1004731x.joss.201702006

Abstract

Abstract: Although GPU has developed rapidly, the core still uses rasterization and it is a hard problem for the complex scene. The ray tracing can simulate the geometrical optics and get the better effect, but it needs the amounts of computing power. Thus, the characteristic of the ray tracing was studied and the new pipeline was designed. A dynamic reconfigurable hardware architecture (RTGPU) was put forward mapping the pipeline to it. A simulation platform was designed. According to results of experiments, more than 11x on RTGPU could be gained achieving the better performance than GPU.

Key words: ray tracing, reconfigurable, pipeline, hardware architecture, simulation

CLC Number:

TP391.9

Cao Xiaopeng, Han Jungang. Reconfigurable RTGPU's Architecture and Simulation[J]. Journal of System Simulation, 2017, 29(2): 273-281.

References

[1] Whitted T.An improved illumination model for shaded display[J]. Communications of the ACM (S0001-0782), 1980, 23(6): 343-349.
[2] Arvo J, Kirk D.Fast ray tracing by ray classification[J]. ACM Siggraph Computer Graphics (S0097-8930), 1987, 21(4): 55-64.
[3] Kang Y, Nah J, Park W, et al.gkDtree: A group-based parallel update kd-tree for interactive ray tracing[J]. Journal of Systems Architecture (S1383-7621), 2013, 59(3): 166-175.
[4] Novák J, Dachsbacher C.Rasterized Bounding Volume Hierarchies[J]. Computer Graphics Forum (S1467-8659), 2012, 31(2): 403-412.
[5] Kalojanov J, Billeter M, Slusallek P.Two-Level Grids for Ray Tracing on GPUs[J]. Computer Graphics Forum (S1467-8659), 2011, 30(2): 307-314.
[6] Schmittler J, Wald I, Slusallek P.SaarCOR: a hardware architecture for ray tracing[C]// Proceedings of Eurographics Workshop on Graphics Hardware. Switzerland: Eurographics Association Aire-la-Ville, 2002: 27-36.
[7] Nery A S, Nedjah N, França F M G. An efficient parallel architecture for ray-tracing[J]. Analog Integrated Circuits & Signal Processing (S0925-1030), 2012, 70(2): 189-202.
[8] Parker S G, Bigler J, Dietrich A, et al.OptiX: A General Purpose Ray Tracing Engine[J]. ACM Transactions on Graphics (S0730-0301), 2010, 29(4): 157-166.
[9] Hu W, Huang Y, Zhang F, et al.Ray tracing via GPU rasterization[J]. Visual Computer (S0178-2789), 2014, 30(6/8): 697-706.
[10] Chochlík M.Scalable multi-GPU cloud raytracing with OpenGL[C]// International Conference on Digital Technologies. USA: IEEE, 2014: 87-95.
[11] Zhou K, Hou Q, Wang R, et al.Real-time kd-tree construction on graphics hardware[J]. ACM Transactions on Graphics (S0730-0301), 2013, 27(5): 126-138.
[12] Zhou K, K Z, Zhou K, et al. Ray tracing on graphics hardware using kd-trees [P]. USA: US20100079451 A1, 2008.
[13] 邹华, 高新波, 吕新荣. 层次包围盒与GPU实现相结合的光线投射算法[J]. 计算机辅助设计与图形学学报(S1003-9775), 2009, 21(2): 172-178.
(Zou Hua, Gao Xin-bo, Lü Xin-rong.A Ray Casting Algorithm Based on Hierarchical Bounding Volumes and GPU[J]. Journal of Computer-Aided Design and Computer Graphics (S1003-9775), 2009, 21(2): 172-178. )
[14] 卢贺齐, 鲍鹏, 冯结青. 基于OpenCL的实时KD—Tree与动态场景光线跟踪[J]. 计算机辅助设计与图形学学报(S1003-9775), 2013, 25(7): 963-973.
(Lu He-qi, Bao Peng, Feng Jieqing.OpenCL Based Real-Time KD-Tree and Ray tracing for Dynamic Scene[J]. Journal of Computer-Aided Design and Computer Graphics (S1003-9775), 2013, 25(7): 963-973.)
[15] Möller T, Trumbore B.Fast, minimum storage ray triangle intersection[J]. Journal of Graphics Tools (S2165-347X), 2005, 2(1): 21-28.
[16] Cao X,Huang H,Li T.A cycle accurate simulation platform for GPU performance study[J]. International Journal of Advancements in Computing Technology (S2005-8039),2012, 4(20): 684-691.
[17] 龚兴全, 李康, 孔凡敏. 基于OpenCL的图形处理器FDTD算法仿真研究[J]. 系统仿真学报(S1004-731X), 2014, 26(8): 1639-1643.
(Gong Xingquan, Li Kang, Kong Fanmin.FDTD Simulation Using Graphic Processing Units Based on OpenCL[J]. Journal of System Simulation (S1004-731X), 2014, 26(8): 1639-1643.)