Crate float16 

A crate that provides support for half-precision 16-bit floating point types.

This crate provides the f16 type, which is an implementation of the IEEE 754-2008 standard binary16 a.k.a “half” floating point type. This 16-bit floating point type is intended for efficient storage where the full range and precision of a larger floating point value is not required. This is especially useful for image storage formats.

This crate also provides a bf16 type, an alternative 16-bit floating point format. The bfloat16 format is a truncated IEEE 754 standard binary32 float that preserves the exponent to allow the same range as f32 but with only 8 bits of precision (instead of 11 bits for f16). See the bf16 type for details.

Because f16 and bf16 are primarily for efficient storage, floating point operations such as addition, multiplication, etc. are not always implemented by hardware. When hardware does not support these operations, this crate emulates them by converting the value to f32 before performing the operation and then back afterward.

Note that conversion from f32/f64 to both f16 and bf16 are lossy operations, and just as converting a f64 to f32 is lossy and does not have Into/From trait implementations, so too do these smaller types not have those trait implementations either. Instead, use from_f32/from_f64 functions for the types in this crate.

The crate supports #[no_std] when the std cargo feature is not enabled, so can be used in embedded environments without using the Rust std library. The std feature enables support for the standard library and is enabled by default, see the Cargo Features section below.

Hardware support

Hardware support for these conversions and arithmetic will be used whenever hardware support is available—either through instrinsics or targeted assembly—although a nightly Rust toolchain may be required for some hardware. When hardware supports it the functions and traits HalfBitsSliceExt and HalfFloatSliceExt are used it will also use vectorized SIMD intructions for increased efficiency.

The following list details hardware support for floating point types in this crate. When using std cargo feature, runtime CPU target detection will be used. To get the most performance benefits, compile for specific CPU features which avoids the runtime overhead and works in a no_std environment.

Architecture	CPU Target Feature	Notes
x86/x86_64	f16c	This supports conversion to/from f16 only (including vector SIMD) and does not support any bf16 or arithmetic operations.
aarch64	fp16	This supports all operations on f16 only.
loongarch64	lsx	This supports conversion to/from f16 only (including vector SIMD) and does not support any bf16 or arithmetic operations.

Cargo Features

To support numerous features, use the [float16-ext] package, which implements its own f16 and bf16 types that support features like serde serializing, zero-copy logic, and more.

Structs

Bf16

A 16-bit floating point type implementing the bfloat16 format.

F16

A 16-bit floating point type implementing the IEEE 754-2008 standard binary16 a.k.a “half” format.

TryFromFloatError

The error type returned when a checked integral type conversion fails.

bf16

A 16-bit floating point type implementing the bfloat16 format.

f16

A 16-bit floating point type implementing the IEEE 754-2008 standard binary16 a.k.a “half” format.

Traits

HalfBitsSliceExt

Extensions to [u16] slices to support reinterpret operations.

HalfFloatSliceExt

Extensions to [f16] and [bf16] slices to support conversion and reinterpret operations.