Collection of utility methods & structs for working with the CUDA language.
Main types:
GpuVec
is the equivalent of aVec
, but is stored on the GPU.Tensor
is an n-dimensional array type that can be easily transferred to and from the GPU.
The main purpose of cuda-macros
is to define three attributes that perform the same
function as the __global__
, __device__
and __host__
attributes do in the C
language.
To use these macros, a build.rs
file must be defined that runs the build()
function
in the cuda-macros-build
crate. See the cuda-macros-test
crate for an example crate.
Attribute(s) | CPU callable | GPU callable | Runs on |
---|---|---|---|
#[global] |
✔️ | ✔️️️️️️ | GPU |
#[device] |
❌ | ️️️✔️️️️️️ | GPU |
#[host] |
✔️ | ❌ | CPU |
The #[host]
and #[device]
attributes can be combined, in which case two versions will be generated, one for the host and one for the device.
The function will be run on whichever system it was called from.
In CUDA terminology, the CPU is the host, and the GPU is the device.