Calculating the minimum distance between triangles on SIMD Hardware.

Calculating the minimum distance between triangles on SIMD Hardware.
复制标题

在 SIMD 硬件上计算三角形之间的最小距离。

DOI:
--
复制
发表时间:
2015
期刊:
影响因子:
--
通讯作者:
T. Koziara
T. Koziara
中科院分区:
--
文献类型:
--
作者:
Konstantinos Krestenitis;T. Koziara

文献摘要

被引文献

相似文献

研究了三维三角形间最小距离的计算问题 SIMD硬件。我们将问题表述为约束最小化,并应用拟牛顿法 方法,同时解决其多个实例在单独的SIMD向量通道。我们测试英特尔的 开源ISPC编译器[0],它提供了一个显式的编程模型,用于编写向量化的 现代CPU和GPU的代码。在目前的沟通中,我们分享了我们的初步经验, 产生一个优化的ISPC实现,并显示一些初步的数值结果。
This paper presents an investigation into calculating the minimum distance between two 3D triangles on SIMD hardware. We formulate the problem as constrained minimization and apply a quasi-Newton method to solve its multiple instances simultaneously on separate SIMD vector lanes. We test Intel’s open-source ISPC compiler [0], which provides an explicit programming model for writting vectorized code for modern CPUs and GPUs. In the current communication we share our initial experience of producing an optimized ISPC implementation and show some preliminary numerical results.