Uni20
Asynchronous tensor and runtime library.
Loading...
Searching...
No Matches
linalg Directory Reference
Directory dependency graph for linalg:

Directories

 async
 
 backends
 
 blas
 
 cpu
 
 cublas
 
 ops
 

Files

 async.hpp
 Opt-in asynchronous Tensor wrappers for dense linear algebra operations.
 
 backend_selector.hpp
 Ordered backend selector values shared by tensor storage and linalg dispatch.
 
 contraction_axes.hpp
 Normalized axis metadata for fixed-rank pairwise tensor contractions.
 
 contraction_strides.hpp
 Stride grouping and direct-GEMM projections for tensor contraction.
 
 decomposition_precision.hpp
 
 dispatch.hpp
 Operation-tag backend dispatch over tensor-view operands.
 
 dispatch_diagnostics.cpp
 
 dispatch_diagnostics.hpp
 Opt-in structured diagnostics for kernel-dispatch walks.
 
 dispatch_error.hpp
 Structured kernel-dispatch failure data.
 
 dispatch_error_presentation.hpp
 Presentation adapter for structured kernel-dispatch failures.
 
 elementwise_functions.hpp
 Named backend-neutral function objects for elementwise transforms.
 
 kernel_attempt.hpp
 Allocation-free runtime results for kernel backend attempts.
 
 least_squares_info.hpp
 
 linalg.hpp
 
 lu_info.hpp
 
 matrix_product_shape.hpp
 Backend-independent dense matrix-product shape validation.
 
 operation_scalar.hpp
 Numeric coefficient sources shared by synchronous and asynchronous operations.
 
 operation_tags.hpp
 Backend-independent operation values used with normalized kernel operands.
 
 precision.hpp
 
 reduction_axes.hpp
 Normalized input and surviving axes for fixed-rank tensor reductions.
 
 refinement_info.hpp
 
 solve_info.hpp
 
 svd_least_squares_info.hpp