IsFinite

Applicability

Product	Supported
Atlas 350 Accelerator Card	√
Atlas A3 training product/Atlas A3 inference product	x
Atlas A2 training product/Atlas A2 inference product	x
Atlas 200I/500 A2 inference product	x
Atlas inference product AI Core	x
Atlas inference product Vector Core	x
Atlas training product	x

Function Usage

Checks whether the input floating-point number is a non-NAN or non-±INF value. The output is a floating-point number or Boolean value. For non-NAN or non-±INF input, if the output is of the floating-point type, the result of the corresponding position is 1 of the floating-point type. If not, the result is 0. If the output is of the Boolean type, the result of the corresponding position is true. If not, the result is false. The formula is as follows:

$\text{[math]}$

If the output is of the floating-point type:
$\text{[math]}$

If the output is of the bool type:
$\text{[math]}$

Prototype

template<typename T, typename U>
__aicore__ inline void IsFinite(const LocalTensor<U>& dst, const LocalTensor<T>& src, uint32_t calCount)

Parameters

**Table 1** Template parameters
Parameter	Description
T	Data type of the source operand. For the Atlas 350 Accelerator Card, the supported data types are half, bfloat16_t, and float.
U	Data type of the destination operand. For the Atlas 350 Accelerator Card, the supported data types are bool, half, bfloat16_t, and float.

**Table 2** API parameters
Parameter	Input/Output	Description
dst	Output	Destination operand. The type is LocalTensor, and TPosition can be VECIN, VECCALC, or VECOUT. The data type of the destination operand is the same as that of the source operand, or the data type of the destination operand is bool. For details about the supported data type groups, see Table 3.
src	Input	Source operand. The type is LocalTensor, and TPosition can be VECIN, VECCALC, or VECOUT.
calCount	Input	Number of elements involved in the computation.

**Table 3** Data type groups supported by the input and output.
srcDtype	dstDtype
half	half
half	bool
float	float
float	bool
bfloat16_t	bfloat16_t
bfloat16_t	bool

Returns

None

Constraints

The source operand address must not overlap the destination operand address.
For details about the alignment requirements of the operand address offset, see General Description and Restrictions.

Examples

For details about the complete operator example, see isfinite operator sample.

// dstLocal: tensor for storing the IsFinite computation result
// srcLocal: input tensor for storing data for IsFinite computation
// The number of elements involved in the computation is 8.
AscendC::IsFinite(dstLocal, srcLocal, 8);

Result example:

The input data type is float, and the output data type is bool.
Input (src):
[1.0,+inf,3.0,4.0,nan,6.0,-inf,8.0]
Output (dst):
[true,false,true,true,false,true,false,true]

Parent topic: IsFinite APIs