arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1404.0075cs.CC
收藏

光线追踪--计算不可计算的东西?

Ray tracing -- computing the incomputable?

Ed Blakey

摘要:

我们从以前的工作中回顾了一个独立于模型的计算复杂性理论框架。值得注意的是,对于本文件,该框架允许形式化的精度问题,目前自己当一个考虑物理的,容易出错的(特别是模拟,而不是数字)计算系统。我们以光线跟踪问题为例,这是一个图灵机无法计算的问题,尽管明显违反了丘奇-图灵命题,但可以说它可以由某些光学计算机解决;然而,我们应用复杂性理论的框架,以便形式化直觉,即一旦适当考虑精度,这些计算机的所谓超级图灵能力实际上就消失了。

英文摘要:

We recall from previous work a model-independent framework of computational complexity theory. Notably for the present paper, the framework allows formalization of the issues of precision that present themselves when one considers physical, error-prone (especially analogue rather than digital) computational systems. We take as a case study the ray-tracing problem, a Turing-machine-incomputable problem that can, in apparent violation of the Church-Turing thesis, nonetheless be said to be solved by certain optical computers; however, we apply the framework of complexity theory so as to formalize the intuition that the purported super-Turing power of these computers in fact vanishes once precision is properly considered.

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。

↑