Edit

Concurrency::fast_math Namespace

  • Article

Functions in the fast_math namespace have lower accuracy, support only single-precision (float), and call the DirectX intrinsics. There are two versions of each function, for example cos and cosf. Both versions take and return a float, but each calls the same DirectX intrinsic.

Syntax

namespace fast_math;

Members

Functions

Name Description
cos Calculates the arccosine of the argument
cosf Calculates the arccosine of the argument
asin Calculates the arcsine of the argument
asinf Calculates the arcsine of the argument
atan Calculates the arctangent of the argument
atan2 Calculates the arctangent of _Y/_X
atan2f Calculates the arctangent of _Y/_X
atanf Calculates the arctangent of the argument
ceil Calculates the ceiling of the argument
ceilf Calculates the ceiling of the argument
cos Calculates the cosine of the argument
cosf Calculates the cosine of the argument
cosh Calculates the hyperbolic cosine value of the argument
coshf Calculates the hyperbolic cosine value of the argument
exp Calculates the base-e exponential of the argument
exp2 Calculates the base-2 exponential of the argument
exp2f Calculates the base-2 exponential of the argument
expf Calculates the base-e exponential of the argument
fabs Returns the absolute value of the argument
fabsf Returns the absolute value of the argument
floor Calculates the floor of the argument
floorf Calculates the floor of the argument
fmax Determine the maximum numeric value of the arguments
fmaxf Determine the maximum numeric value of the arguments
fmin Determine the minimum numeric value of the arguments
fminf Determine the minimum numeric value of the arguments
fmod Calculates the floating-point remainder of _X/_Y
fmodf Calculates the floating-point remainder of _X/_Y
frexp Gets the mantissa and exponent of _X
frexpf Gets the mantissa and exponent of _X
isfinite Determines whether the argument has a finite value
isinf Determines whether the argument is an infinity
isnan Determines whether the argument is a NaN
ldexp Computes a real number from the mantissa and exponent
ldexpf Computes a real number from the mantissa and exponent
log Calculates the base-e logarithm of the argument
log10 Calculates the base-10 logarithm of the argument
log10f Calculates the base-10 logarithm of the argument
log2 Calculates the base-2 logarithm of the argument
log2f Calculates the base-2 logarithm of the argument
logf Calculates the base-e logarithm of the argument
modf Splits _X into fractional and integer parts.
modff Splits _X into fractional and integer parts.
pow Calculates _X raised to the power of _Y
powf Calculates _X raised to the power of _Y
round Rounds _X to the nearest integer
roundf Rounds _X to the nearest integer
rsqrt Returns the reciprocal of the square root of the argument
rsqrtf Returns the reciprocal of the square root of the argument
signbit Returns the sign of the argument
signbitf Returns the sign of the argument
sin Calculates the sine value of the argument
sincos Calculates sine and cosine value of _X
sincosf Calculates sine and cosine value of _X
sinf Calculates the sine value of the argument
sinh Calculates the hyperbolic sine value of the argument
sinhf Calculates the hyperbolic sine value of the argument
sqrt Calculates the square root of the argument
sqrtf Calculates the square root of the argument
tan Calculates the tangent value of the argument
tanf Calculates the tangent value of the argument
tanh Calculates the hyperbolic tangent value of the argument
tanhf Calculates the hyperbolic tangent value of the argument
trunc Truncates the argument to the integer component
truncf Truncates the argument to the integer component

Requirements

Header: amp_math.h

Namespace: Concurrency::fast_math

See also

Concurrency Namespace (C++ AMP)