Module cuda_backend 

Real CUDA GPU backend using AvarokRegistry.

SBIO IORouter: all CUDA operations flow through GpuBackend. Uses AvarokRegistry for kernel loading/launching and raw CUDA driver API for memory management.

Modules

arch_preflight

Refuse to load kernels the GPU cannot run, BEFORE the driver does it badly.

tensormap

TMA descriptors (CUtensorMap) for cp.async.bulk.tensor loads.

Structs

AvarokCudaBackend

Production GPU backend wrapping AvarokRegistry + raw CUDA driver API.

Functions

cuda_free_memory_bytes

Query GPU free memory without requiring a GpuBackend reference. Safe to call from any thread that shares the CUDA context.

spawn_oom_watchdog

Start the OOM watchdog, or return None if one is already running.