Module linalg_dispatch 

Automatic GPU dispatch shim for dense linear algebra hot kernels.

Every try_* entry point in this module is invoked unconditionally from gam_linalg::faer_ndarray before the CPU fast-path runs. The decision to send the kernel to a device is fully automatic and never requires a user-facing flag — it depends only on:

1. GpuRuntime::global() returning Some(_) (a device was probed at process startup).
2. The kernel being large enough to amortize launch/PCIe overhead, per the thresholds in policy::GpuDispatchPolicy.
3. cudarc successfully dynamically loading libcuda at process startup via its fallback-dynamic-loading feature. When the loader fails (no driver, no toolkit installed), GpuRuntime::probe() returns Ok(None) and every try_* returns None so the caller falls through to the existing faer CPU kernel.

The wiring lives here so solver/pirls.rs and the family Hessian assemblers can stay backend-agnostic: they call gam_linalg::faer_ndarray::fast_* and get GPU acceleration automatically whenever it is profitable.

Structs

CudaGemmDispatch

ResidentDesignGram

#1017 Phase 3: a device-resident design matrix for repeated Xᵀ·diag(w)·X Gram evaluations that uploads X to the device ONCE.

Enums

DispatchOp

Discriminator used by route_through_gpu to apply the right size threshold from super::policy::GpuDispatchPolicy.

Functions

route_through_gpu

Returns Some(runtime) when both a device is available and the workload is large enough per policy. The caller can then attempt the actual device kernel; any backend failure is expected to return None from the lower layer and the CPU fast path resumes.

try_cholesky_batched_lower_inplace

try_cholesky_lower_inplace

try_fast_ab

try_fast_ab_broadcast_b_batched

try_fast_abt_strided_batched

try_fast_atb

try_fast_atb_on_ordinal

Aᵀ·B on a specific device ordinal, for pool-tiled callers that already own the ordinal (the worker thread has bound that ordinal’s context). Semantics are identical to try_fast_atb — a is m×k, b is m×n, output is the k×n product aᵀ·b — but the kernel is pinned to ordinal instead of the probe-selected primary device. Returns None when CUDA is unavailable, the shape is below policy threshold, or the backend reports a transient failure, so the caller runs its CPU fallback. f64 only.

try_fast_atv

try_fast_av

try_fast_joint_hessian_2x2

try_fast_spectral_leverage_diagonal

GPU-offloaded spectral leverage diagonal h[i] = ‖(X G)_{i,:}‖².

try_fast_xt_diag_x

try_fast_xt_diag_y

try_solve_lower_triangular_matrix

try_solve_upper_triangular_matrix