## C Specification

#pragma OPENCL EXTENSION cl_khr_fp16 : enable

## Description

This extension adds support for half scalar and vector types as built-in types that can be used for arithmetic operations, conversions, etc. An application that wants to use half and halfn types will need to include the directive shown above.

The list of built-in scalar and vector data types are extended to include the types in the table below.

The built-in vector data types for halfn are also declared as appropriate types in the OpenCL API (and header files) that can be used by an application. The following table describes the built-in vector data types for halfn as defined in the OpenCL C programming language and the corresponding data type available to the application:

Type in OpenCL Language Description API type for application

half2

2-component half-precision floating-point vector

cl_half2

half3

3-component half-precision floating-point vector

cl_half3

half4

4-component half-precision floating-point vector

cl_half4

half8

8-component half-precision floating-point vector

cl_half8

half16

16-component half-precision floating-point vector

cl_half16

The relational, equality, logical and logical unary operators can be used with half scalar and halfn vector types and shall produce a scalar int and vector shortn result respectively.

The OpenCL compiler accepts an h and H suffix on floating point literals, indicating the literal is typed as a half.

The macro names given in the following list must use the values specified. These constant expressions are suitable for use in #if preprocessing directives.

The following table also describes the corresponding macro names available to the application.

Macro in OpenCL Language value Macro for application

#define HALF_DIG

3

HALF_DIG

#define HALF_MANT_DIG

11

HALF_MANT_DIG

#define HALF_MAX_10_EXP

+4

HALF_MAX_10_EXP

#define HALF_MAX_EXP

+16

HALF_MAX_EXP

#define HALF_MIN_10_EXP

-4

HALF_MIN_10_EXP

#define HALF_MIN_EXP

-13

HALF_MIN_EXP

#define HALF_RADIX

2

HALF_RADIX

#define HALF_MAX

0x1.ffcp15h

HALF_MAX

#define HALF_MIN

0x1.0p-14h

HALF_MIN

#define HALF_EPSILON

0x1.0p-10h

HALF_EPSILON

The following constants are also available. They are of type half and are accurate within the precision of the half type.

Constant Description

M_E_H

Value of e

M_LOG2E_H

Value of log2 e

M_LOG10E_H

Value of log10 e

M_LN2_H

Value of ln 2

M_LN10_H

Value of ln 10

M_PI_H

Value of π

M_PI_2_H

Value of π / 2

M_PI_4_H

Value of π / 4

M_1_PI_H

Value of 1 / π

M_2_PI_H

Value of 2 / π

M_2_SQRTPI_H

Value of 2 / √π

M_SQRT2_H

Value of √2

M_SQRT1_2_H

Value of 1 / √2

Ann application may query the configuration information using the op-code CL_DEVICE_HALF_FP_CONFIG with clGetDeviceInfo for an OpenCL device that supports half precision floating-point.

Conversions

The implicit conversion rules specified in section 6.2.1 now include the half scalar and halfn vector data types.

The explicit casts described in section 6.2.2 are extended to take a half scalar data type and a halfn vector data type.

The explicit conversion functions described in section 6.2.3 are extended to take a half scalar data type and a halfn vector data type.

The as_typen() function for re-interpreting types as described in section 6.2.4.2 is extended to allow conversion-free casts between shortn, ushortn and halfn scalar and vector data types.