simdpp::max(floatint-point)

 

libsimdpp

Types

Capabilities

Operations

Conversions

Bitwise operations

Floating-point operations

Integer operations

Memory access operations

Shuffle operations

Miscellaneous operations

 

Floating-point operations

Elementary operations
add sub	mul div	fmadd fmsub	neg
Rounding
round ceil	floor trunc
Comparisons
cmp_eq cmp_neq	cmp_lt cmp_le	cmp_gt cmp_ge
Reduction
reduce_add reduce_min	reduce_mul reduce_max
Miscellaneous
abs sign	min max	isnan isnan2
Square root and reciprocal
sqrt	rcp_e rcp_rh	rsqrt_e rsqrt_rh

 

template<unsigned N, class E1, class E2> Ret<N, _DETAIL_> max( Vec1<N,E1> a, Vec2<N,E2> b );

Computes the minimum of the values in two floating-point values. The implementation behaves as if the following set of overloads is provided:

Ret	Vec1	Vec2
float32	float32	float32
float64	float64	float64

The return type is a vector expression.

For each pair of values being processed: if at least one of the values is NaN, or both values are zeroes, it is unspecified which value will be returned.

Scalar arguments

Several additional overloads are provided to implement implicit conversions from scalar types. Values of the following types are accepted as the first or the second argument: int32_t, int64_t, uint32_t, uint64_t, float, double. Each element of the expanded vector is set to the source value, converting it using standard conversions if needed.

The return type is a vector expression. The vector type is the same as the vector argument except that masks are unmasked to respective non-mask vector types.

Parameters

a, b

-

floating-point vectors

Return value

A vector expression evaluating to max(a, b).

Equivalent operation

r0 = a0 > b0 ? a0 : b0
...
rN = aN > bN ? aN : bN