Enum LayoutSku 

pub enum LayoutSku {
    Rcr,
    Rrr,
}

Layout SKU. Describes the row/column orientation of A, B, C, and D for matrix-multiply-shaped kernels.

Intentionally NOT #[non_exhaustive] — the GEMM layout space is essentially closed in practice (row-major / column-major permutations of A, B, C/D); the two wired variants cover the dispatch space baracuda-cutlass selects against. New variants would be a deliberate breaking change with a major-version bump.

Variants

Rcr

A row-major [M, K], B column-major [K, N], C/D row-major [M, N].

Useful when a row-major weight tensor stored as [N, K] is reinterpreted as logical column-major B = [K, N] without a transpose copy.

Rrr

A row-major [M, K], B row-major [K, N], C/D row-major [M, N].

The natural shape for activation-row-major @ weight-row-major matmul (the typical ML graph layout). No transpose pass needed before launch — both operands stored in their native row-major form.

Trait Implementations

impl Clone for LayoutSku

fn clone(&self) -> LayoutSku

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Copy for LayoutSku

impl Debug for LayoutSku

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.

impl Eq for LayoutSku

impl Hash for LayoutSku

fn hash<__H>(&self, state: &mut __H)

where __H: Hasher,

Feeds this value into the given Hasher.

fn hash_slice<H>(data: &[Self], state: &mut H)

where H: Hasher, Self: Sized,

Feeds a slice of this type into the given Hasher.

impl PartialEq for LayoutSku

fn eq(&self, other: &LayoutSku) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl StructuralPartialEq for LayoutSku