(图 via WCCFTech)
出于好奇,他用 VS Code 打开了该文件。毕竟对于常规的记事本(notepad.exe)应用程序来说,该 txt 文档的体型着实大得有些离谱。
结果在这个包含 65340 行的文档中,首先映入眼帘的就是 —— 列夫·托尔斯泰的古腾堡计划,《战争与和平》电子书。
尽管让人感到有些疑惑,还是有网友给出了相对合理的解释。据悉,作为计算平台和编程模型的 NVIDIA CUDA,开发者可在该公司的 GPU 上开展一系列通用加速计算。
然后 Jlouis8 在评论中贴出了 GitHub 上的一个代码示例(传送门),演示了如何通过创建一个小型的 CUDA 内核,以计算某些数据(比如《战争与和平》电子书)中,w、x、y 和 z 的字母数量。
之所以选择这本电子书,其原因也不难理解。首先,NVIDIA 可在不侵犯版权的情况下自由包含;其次,原始文本量足够庞大、但又足够在 CUDA(GPU)硬件上快速完成。
(figure via WCCFTech)
Out of curiosity, he opened the file with VS Code. After all, for a regular notepad.exe application, the txt document is ridiculously large.
As a result, the first thing to see in this 65340-line document is Leo Tolstoy’s Gutenberg Project, the ebook War and Peace.
Although there is some confusion, some netizens have given a relatively reasonable explanation. It is reported that NVIDIA CUDA, as a computing platform and programming model, developers can carry out a series of general accelerated computing on the company’s GPU.
In the comments, Jlouis8 then posted a code example (portal) on GitHub, demonstrating how to create a small CUDA kernel to count the letters w, x, y, and z in some data, such as the war and peace ebook.
The reason for choosing this e-book is not difficult to understand. First, NVIDIA can be included freely without copyright infringement; second, the amount of original text is large enough, but fast enough to be done on CUDA (GPU) hardware.