Every story and comment from the Hacker News API, in a layout borrowed from 2001.
A comment on Show HN: Flash Attention in ~100 lines of CUDA. Parent comment
There's also explicit warp synchronization, i.e. __syncwarp(). More on warp primitives here: https://developer.nvidia.com/blog/using-cuda-warp-level-prim...
There are no comments on this comment yet.
There's also explicit warp synchronization, i.e. __syncwarp(). More on warp primitives here: https://developer.nvidia.com/blog/using-cuda-warp-level-prim...