Floating-Point Options

This section contains descriptions for compiler options that pertain to floating-point calculations.

• fast-transcendentals, Qfast-transcendentals
  Enables the compiler to replace calls to transcendental functions with faster but less precise implementations.
• fimf-absolute-error, Qimf-absolute-error
  Defines the maximum allowable absolute error for math library function results.
• fimf-accuracy-bits, Qimf-accuracy-bits
  Defines the relative error for math library function results, including division and square root.
• fimf-arch-consistency, Qimf-arch-consistency
  Ensures that the math library functions produce consistent results across different microarchitectural implementations of the same architecture.
• fimf-domain-exclusion, Qimf-domain-exclusion
  Indicates the input arguments domain on which math functions must provide correct results.
• fimf-force-dynamic-target, Qimf-force-dynamic-target
  Instructs the compiler to use run-time dispatch in calls to math functions.
• fimf-max-error, Qimf-max-error
  Defines the maximum allowable relative error for math library function results, including division and square root.
• fimf-precision, Qimf-precision
  Lets you specify a level of accuracy (precision) that the compiler should use when determining which math library functions to use.
• fimf-use-svml, Qimf-use-svml
  Instructs the compiler to use the Short Vector Math Library (SVML) rather than the Intel® C++ Compiler Classic Math Library (LIBM) to implement math library functions.
• fma, Qfma
  Determines whether the compiler generates fused multiply-add (FMA) instructions if such instructions exist on the target processor.
• fp-model, fp
  Controls the semantics of floating-point calculations.
• fp-port, Qfp-port
  Rounds floating-point results after floating-point operations.
• fp-speculation, Qfp-speculation
  Tells the compiler the mode in which to speculate on floating-point operations.
• fp-stack-check, Qfp-stack-check
  Tells the compiler to generate extra code after every function call to ensure that the floating-point stack is in the expected state.
• fp-trap, Qfp-trap
  Sets the floating-point trapping mode for the main routine.
• fp-trap-all, Qfp-trap-all
  Sets the floating-point trapping mode for all routines.
• ftz, Qftz
  Flushes denormal results to zero.
• Ge
  Enables stack-checking for all functions. This is a deprecated option. The replacement option is /Gs0.
• mp1, Qprec
  Improves floating-point precision and consistency.
• pc, Qpc
  Enables control of floating-point significand precision.
• prec-div, Qprec-div
  Improves precision of floating-point divides.
• prec-sqrt, Qprec-sqrt
  Improves precision of square root implementations.
• qsimd-honor-fp-model, Qsimd-honor-fp-model
  Tells the compiler to obey the selected floating-point model when vectorizing SIMD loops.
• qsimd-serialize-fp-reduction, Qsimd-serialize-fp-reduction
  Tells the compiler to serialize floating-point reduction when vectorizing SIMD loops.
• rcd, Qrcd
  Enables fast float-to-integer conversions. This is a deprecated option. There is no replacement option.