博客：
基于CUDA的并行程序设计，https://blog.csdn.net/manong_wxd/article/category/7303563
CUDA学习笔记，https://blog.csdn.net/qq_36387683/article/category/7819160
CUDA从入门到精通，https://blog.csdn.net/LG1259156776/article/details/52491431
CUDA框架介绍，https://blog.csdn.net/LG1259156776/article/details/52491438
__syncthreads的理解，https://blog.csdn.net/LG1259156776/article/details/52684868
显卡里的cuda真正做到并行运算的机制和线程个数，主要关于SM、warp等概念，https://blog.csdn.net/yangdashi888/article/details/78183464/
NVIDIA显卡计算能力对照比，https://developer.nvidia.com/cuda-gpus