atomr-accel-cuda 0.10.0

GPU acceleration via the actor model. Wraps NVIDIA CUDA libraries (cuBLAS, cuDNN, cuFFT, cuRAND, cuSOLVER, cuSPARSE, cuTENSOR, cuBLASLt, NVRTC, NCCL) as supervised atomr actors with generation-validated buffers and a uniform async surface.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19

//! `DeviceActor` (outer tier) + `ContextActor` (inner tier) — §5.11.

mod alloc_dispatch;
mod alloc_msg;
mod context_actor;
pub mod device_actor;
mod state;

pub use alloc_dispatch::{
    AllocDispatch, AllocReq, CopyFromHostDispatch, CopyFromHostReq, CopyToHostDispatch,
    CopyToHostReq,
};
pub use alloc_msg::{DeviceLoad, HostBuf};
pub use context_actor::{ContextActor, ContextMsg};
pub use device_actor::{
    DeviceActor, DeviceConfig, DeviceMsg, EnabledLibraries, KernelChildren, SgemmRequest,
    WorkRequest,
};
pub use state::DeviceState;