QuietNaN

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

Returns a quiet NaN value of a specified data type (the most significant bit of the floating-point mantissa is 1).

For the Atlas 350 Accelerator Card, the supported data types are half, bfloat16_t, and float.

Scalar API, which returns a scalar.

constexpr __aicore__ static inline T QuietNaN()

Vector API, which assigns a quiet NaN value to the first count elements of dstLocal.

__aicore__ static inline void QuietNaN(const LocalTensor<T>& dstLocal, uint32_t count)

**Table 1** Parameters
Parameter	Input/Output	Description
dstLocal	Output	Destination operand. Parameter of the LocalTensor type. TPosition can be VECIN, VECCALC, or VECOUT. The start address of LocalTensor must be 32-byte aligned.
count	Input	Number of elements of the input data.

The scalar API returns a quiet NaN value of the corresponding data type.

The vector API has no return value.

None

Scalar API

float value = AscendC::NumericLimits<float>::QuietNaN();

Vector API

AscendC::NumericLimits<float>::QuietNaN(dstLocal, 256); // Return a quiet NaN value for 256 elements of float.

Parent topic: NumericLimits