Enum StoragePrecision 

pub enum StoragePrecision {
    Fp16,
    Bf16,
}

Storage precision for matrix values.

Variants

Fp16

IEEE 754 half-precision (FP16, 10-bit mantissa).

Bf16

Brain floating-point (BF16, 7-bit mantissa, wider exponent range).

Implementations

impl StoragePrecision

pub const fn ptx_type(self) -> PtxType

Returns the PTX type corresponding to this storage precision.

pub const fn packed_ptx_type(self) -> PtxType

Returns the packed (x2) PTX type for this precision.

pub const fn element_bytes(self) -> u32

Bytes per element.

pub const fn mantissa_bits(self) -> u32

Mantissa bits (for error analysis).

pub const fn suffix(self) -> &'static str

Returns the PTX type suffix string without the leading dot.

Trait Implementations

impl Clone for StoragePrecision

fn clone(&self) -> StoragePrecision

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for StoragePrecision

impl Debug for StoragePrecision

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

Formats the value using the given formatter.

impl Eq for StoragePrecision

impl Hash for StoragePrecision

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 StoragePrecision

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