MatrixCoord wraps Coord<2, int> to provide a helper for accessing named dimensions. Classes expecting a coordinate in the rank=2 index space of a matr

Coordinate in pitch-linear space

IEEE half-precision floating-point type

Floating-point type with 8 bits of exponent and 7 bits of mantissa.

Tensor Float 32 data type

Object to encapsulate the fast division+modulus operation. This object precomputes two values used to accelerate the computation and is best used whe

GemmCoord is a structure derived from Coord<3> that specifies a location within the coordinate space of a GEMM problem.

Mapping function for pitch-linear memory

Object to encapsulate the fast division+modulus operation for 64b integer division. This object precomputes two values used to accelerate the computa

////////////////////////////////////////////////////////// floating-point 8 type : E4M3 //////////////////////////////////////////////////////////

Forward declaration of float_e5m2_t to define float_e4m3_t <=> float_e5m2_t conversions in class float_e4m3_t

Mask object

Defines a canonical 4D coordinate used by tensor operations.

Defines a canonical 5D coordinate used by tensor operations.

Object to encapsulate the fast division+modulus operation for 64b integer division in which the divisor is a power of two.

BatchedGemmCoord is a structure derived from Coord<4> that specifies a location within the coordinate space of a batched GEMM problem.

! Unsigned 128b integer type

Device-side kernel params Arguments has the untransformed problem shape from the user. Params will have the transformed problem shape.

FP4 and FP6 types

Mapping function for scenario in which both rows and columns are separated by a stride. Row stride is smaller than column stride in AffineRank2ColumnM

Mapping function for scenario in which both rows and columns are separated by a stride. Column stride is smaller than row stride in AffineRank2RowMajo

Parameters object for Conv2d DGRAD Filter (w) iterator

Predicated tile access iterator descriptor object containing template dependent state

Mapping function for 4-D NHWC tensors.

////////////////////////////////////////////////////////// floating-point 8 type : UE4M3 //////////////////////////////////////////////////////////

Bidirectional iterator over elements

Argument structure

Per-cell input shape that backends consume. Fields match the kernel-visible shape (post TP/EP simulation), not the full-model shape. The dri

Mapping function for column-major matrices.

Template based on element size (in bits) - defined in terms of pitch-linear memory and Crosswise size (in elements).

Template mapping a column-major view of pitch-linear memory to TensorOpMultiplicand

Template mapping a column-major view of pitch-linear memory to TensorOpMultiplicand

Template mapping a column-major view of pitch-linear memory to TensorOpMultiplicand