Docs.rs

OxiCUDA Rand - GPU-accelerated random number generation (cuRAND equivalent)

GPU-accelerated training engine: fused optimizer update kernels, gradient checkpointing, LR schedulers, and ZeRO optimizer state sharding

OxiCUDA BLAS - GPU-accelerated BLAS operations (cuBLAS equivalent)

OxiCUDA Primitives - CUB-equivalent parallel GPU primitives (radix sort, scan, reduce, select, histogram)

OxiCUDA Graph — CUDA Graph execution engine with operator fusion, buffer lifetime analysis, stream partitioning, and optimized execution planning

OpenTalk RoomServer Modules

OxiCUDA Autotune - Automatic GPU kernel parameter optimization engine

OpenTalk RoomServer Module Training Participation Report

OxiCUDA Launch - Type-safe GPU kernel launch infrastructure

OxiCUDA Runtime - CUDA Runtime API wrapper (cudaMalloc/cudaMemcpy/cudaLaunchKernel) built on the driver API

OxiCUDA PTX - PTX code generation DSL and IR for GPU kernel development

OxiCUDA Memory - Type-safe GPU memory management with Rust ownership semantics

OpenTalk RoomServer Meeting Notes

OpenTalk RoomServer Module Whiteboard

OxiCUDA Driver - Dynamic CUDA driver API wrapper via libloading (zero SDK dependency)