期刊文献+

并行计算在光线追踪中的应用 被引量:2

Application of Parallel Algorithm in Ray Tracing
在线阅读 下载PDF
导出
摘要 目前,计算机图形学的技术包括有光栅化,光线追踪,辐射度3种主流算法。而光线追踪又以其绘制的真实感性及实现的方便性,受到了广泛的应用。但光线追踪的最大问题在于性能。围绕光线追踪中的核心算法-射线与三角形相交算法展开讨论,引入基于Intel TBB并行编程工具的线程级并行技术以及基于SIMD的指令并行技术,提高其算法的速度。其加速技术可以使光线跟踪的性能在计算机多核的情况下,相对于串行跟踪方法有明显的改善。 Rasterization, Ray tracing, Radiosity are the three kinds of major and popular technologies in Graphics. Especially ray tracing is even more popular today because of its realistic rendering and computability. But the performance is the top one bottleneck in ray tracing computation. This paper introduces two parallel algorithms to accelerate the key part computation of ray tracing - ray intersection with triangle: threading level paralleling (based on TBB) and instruction level paralleling (based on SIMD). The result shows that the accelerating algorithm given by this paper computes faster than the serial ray tracing algorithm under the condition of not reducing the light numbers which take part in tracing.
作者 陈是权
机构地区 上海交通大学
出处 《微型电脑应用》 2012年第9期20-22,共3页 Microcomputer Applications
关键词 并行计算 TBB SIMD 多核 光线追踪 射线三角形求交 Parallel Algorithm TBB SIMD Multi-core Ray Tracing Ray-triangle Intersection
  • 相关文献

参考文献4

  • 1Tomas Nikodym (June). Ray Tracing Algorithm ForInteractive Applications[D]. Czech Technical University,FEE, 2010.
  • 2Keith Morley. RealisticRayTracing [M]. 2009.
  • 3Reinders, James. Intel Threading Building Blocks:Outfitting C++ for Multi-core Processor Parallelism(Paperback) [S]. 2007.
  • 4刘刚,梁晓庚.基于SIMD硬件指令加速的并行光线跟踪算法[J].第十届中国科协年会论文集,2008.

同被引文献19

引证文献2

二级引证文献5

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