cuda-kat
CUDA kernel author's tools
|
Some common definitions for all on-device collaborative sequence operations. More...
Classes | |
struct | kat::collaborative::detail::elements_per_lane_in_full_warp_write< T > |
In a "full warp write", we want [1] each lane to write an integral number native words (at the moment and for the foreseeable future, 4-byte integers). More... | |
Enumerations | |
enum | inclusivity_t : bool { Exclusive = false, Inclusive = true } |
Some common definitions for all on-device collaborative sequence operations.