-fp-model compiler option
   Floating-point Options Quick Reference
   Using the -fp-model (/fp) Option
-ftz compiler option
/fp compiler option
   Floating-point Options Quick Reference
   Using the -fp-model (/fp) Option
/fpe compiler option
/Qftz compiler option
   Setting the FTZ and DAZ Flags
   Using the Floating-Point Exception Handling (/fpe) Compiler Option



architectures
    coding guidelines for
arithmetic exception handling
    /fpe options for floating-point data
array
array operation
avoid
    inefficient data types
    mixed arithmetic expressions



checking
    floating-point stacks
    stacks
complex data type
    native IEEE representation (COMPLEX*16)
    native IEEE representation (COMPLEX*32)
    native IEEE representation (COMPLEX*8)
COMPLEX(KIND=16) representation
COMPLEX(KIND=4) representation
COMPLEX(KIND=8) representation



data file
    limitations in converting unformatted files
data format
    floating-point
data types
    COMPLEX
    DOUBLE COMPLEX
    DOUBLE PRECISION
    efficiency
    IEEE S_float representation (COMPLEX*8)
    IEEE S_float representation (REAL*4)
    IEEE T_float representation (COMPLEX*16)
    IEEE T_float representation (COMPLEX*32)
    IEEE T_float representation (REAL*8)
    native IEEE* floating-point representation
    REAL
DAZ flag
denormal exceptions
denormal numbers
denormalized numbers (IEEE*)
    exponent value of
denormals



efficiency
efficient data types
errors
    loss of precision
    rounding



fields in record structures
floating-point
floating-point array operation
floating-point conversion limitations
floating-point data types
    conversion limitations
    digits of precision for REAL*4
    digits of precision for REAL*8
    IEEE* S_float
    IEEE* T_float
    representation of native formats
floating-point exceptions
    /fpe compiler option
    denormal exceptions
    Fortran console applications
    Fortran DLL applications
    Fortran QuickWin applications
    Fortran Standard Graphics applications
    Fortran Windows applications
    types of
floating-point numbers
    formats for
    functions returning parameters
    loss of precision errors
    overflow errors
    overview of
    representation of
    rounding errors
       Loss of Precision Errors
       Rounding Errors
    special values
    underflow errors
floating-point performance
floating-point representation
floating-point stack
floating divide-by-zero
floating invalid
floating overflow
floating underflow
fordef.for file
Fortran console applications
    floating-point exceptions
Fortran DLL applications
    floating-point exceptions
Fortran QuickWin applications
    floating-point exceptions
Fortran Standard Graphics applications
    floating-point exceptions
Fortran Windows* applications
    floating-point exceptions
FTZ flag



IEEE*
    floating-point formats
    floating-point values
    S_float data
       Native IEEE* Floating-Point Representations Overview
       REAL(KIND=4) (REAL) Representation
    S_float representation (COMPLEX*8)
    T_float data
       Native IEEE* Floating-Point Representations Overview
       REAL(KIND=8) (DOUBLE PRECISION) Representation
    T_float representation (COMPLEX*16)
    T_float representation (COMPLEX*32)
    X_floating format
IEEE* floating-point representation



loss of precision errors



machine epsilon
maintainability
MXCSR register



NaN values
normalized floating-point number
Not-a-Number (NaN)
numeric conversion
    limitations of
numeric parameters
    functions returning
numeric representations
    retrieving parameters of



overflow errors
overview



performance



real data type
    native IEEE* representation
    range for REAL*4
    range for REAL*8
REAL(16)
    representation
REAL(4)
    representation
REAL(8)
    representation
relative errors
rounding errors
    machine epsilon
    magnitude of
    relative
    ULPs
run-time performance
    improving



signed infinity
signed zero
stacks
subnormal numbers



ULPs
underflow errors



warning messages
    floating-point overflow (run-time)
    floating-point underflow (run-time)