arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1610.07394cs.DC
收藏

离散正交单形递归GPU映射的可能性

Possibilities of Recursive GPU Mapping for Discrete Orthogonal Simplices

Cristóbal A. Navarro, Benjamín Bustos, Nancy Hitscheld

摘要:

研究了离散正交$m$ -单形的并行线程映射问题。从并行空间效率和潜在性能改进的角度分析了$O(1)$时间递归块空间映射$λ: \mathbb{Z}^m \mapsto \mathbb{Z}^m$的可能性。$2$ -单纯形和$3$ -单纯形作为特殊情况进行分析,其中发现恒定时间映射,分别提供高达$2\times$和$6\times$的潜在改进,比边界框方法更有效。对于一般情况下,它表明,找到一个有效的递归并行空间的$m$ -单纯形取决于两个参数的选择,提供了一些见解,这可以导致一个体积相匹配的$m$ -单纯形$n>n_0$,使并行空间约$m!$倍更有效的比一个边界框。

英文摘要:

The problem of parallel thread mapping is studied for the case of discrete orthogonal $m$-simplices. The possibility of a $O(1)$ time recursive block-space map $λ: \mathbb{Z}^m \mapsto \mathbb{Z}^m$ is analyzed from the point of view of parallel space efficiency and potential performance improvement. The $2$-simplex and $3$-simplex are analyzed as special cases, where constant time maps are found, providing a potential improvement of up to $2\times$ and $6\times$ more efficient than a bounding-box approach, respectively. For the general case it is shown that finding an efficient recursive parallel space for an $m$-simplex depends of the choice of two parameters, for which some insights are provided which can lead to a volume that matches the $m$-simplex for $n>n_0$, making parallel space approximately $m!$ times more efficient than a bounding-box.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑