Struct FusionLimits 

pub struct FusionLimits {
    pub max_elementwise_steps: u32,
    pub max_elementwise_inputs: u32,
}

Hardware / encoder limits for fusion passes.

Fields

max_elementwise_steps: u32

Max steps in one Op::ElementwiseRegion chain (Metal/wgpu/CUDA: 32).

max_elementwise_inputs: u32

Max distinct external inputs in one region (Metal/wgpu/CUDA: 16).

Implementations

impl FusionLimits

pub const GPU_NATIVE: FusionLimits

Caps shared by native elementwise-region kernels today.

pub const UNBOUNDED: FusionLimits

No practical cap — used when regions are unfused to primitives (CPU).

Trait Implementations

impl Clone for FusionLimits

fn clone(&self) -> FusionLimits

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for FusionLimits

impl Debug for FusionLimits

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

Formats the value using the given formatter.

impl Default for FusionLimits

fn default() -> FusionLimits

Returns the “default value” for a type.

impl Eq for FusionLimits

impl PartialEq for FusionLimits

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