HALF Data Type

An alias to uint16_t packed with a 16-bit floating-point number consisting of a sign bit, a 5-bit biased exponent, and a 10-bit mantissa.

typedef uint16_t HALF;


The HALF data type is equivalent to the IEEE 754 binary16 format.

HALF_MIN = 6.10352e-5f

HALF_MAX = 65504.f

For more info about the HALF data type, see Half-precision floating-point format.

Namespace: Use DirectX::PackedVector

Platform Requirements

Microsoft Visual Studio 2010 or Microsoft Visual Studio 2012 with the Windows SDK for Windows 8. Supported for Win32 desktop apps, Windows Store apps, and Windows Phone 8 apps.



See also

DirectXMath Library Types