h2rsqrt

Applicability

Function Usage

Obtains the reciprocal of the square root of each element in the input data x.

Prototype

1

inline half2 h2rsqrt(half2 x)

1

inline bfloat16x2_t h2rsqrt(bfloat16x2_t x)

Parameters

Returns

Reciprocal of the square root of each element in the input data.

• If the input element is 0, the return value is inf.
• If the input element is inf, the return value is 0.
• If the input element is -inf, the return value is nan.
• If the input element is nan, the return value is nan.

Restrictions

The input data x supported by this API must be greater than or equal to 0. Otherwise, the return value is nan.

For SIMT programming, this API is not supported.

Header File to Be Included

To use the half2 API, the simt_api/asc_fp16.h header file must be included. To use the bfloat16x2_t API, the simt_api/asc_bf16.h header file must be included.

1

#include "simt_api/asc_fp16.h"

1

#include "simt_api/asc_bf16.h"

Examples

1
2
3
4
5

__simt_vf__ __launch_bounds__(1024) inline void KernelRsqrt(__gm__ half2* dst, __gm__ half2* x)
{
    int idx = threadIdx.x + blockIdx.x * blockDim.x;
    dst[idx] = h2rsqrt(x[idx]);
}