Enum ConvAlgorithm 

pub enum ConvAlgorithm {
    ImplicitGemm,
    Im2colGemm,
    Winograd,
    Direct,
    FftConv,
}

Convolution algorithm selection.

Different algorithms offer different trade-offs between workspace memory and compute throughput. The optimal choice depends on tensor sizes, GPU architecture, and available workspace.

Variants

ImplicitGemm

Implicit GEMM – low workspace, moderate speed.

Im2colGemm

Im2col followed by explicit GEMM – higher workspace, often fastest for medium-sized feature maps.

Winograd

Winograd transform – fastest for 3x3 kernels with stride 1, but requires workspace and may reduce numerical precision.

Direct

Direct convolution – no workspace, straightforward nested loops.

FftConv

FFT-based convolution – fastest for very large kernels.

Trait Implementations

impl Clone for ConvAlgorithm

fn clone(&self) -> ConvAlgorithm

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for ConvAlgorithm

impl Debug for ConvAlgorithm

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

Formats the value using the given formatter.

impl Eq for ConvAlgorithm

impl Hash for ConvAlgorithm

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

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 ConvAlgorithm

fn eq(&self, other: &ConvAlgorithm) -> 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 ConvAlgorithm