SPEC Accel OpenMP Flag Description for the Intel(R) C/C++ Compiler for IA32 and Intel 64 applications and Intel(R) Fortran Compiler for IA32 and Intel 64 applications

Adds the directory for include files to the search path at compile time.

Adds the directory for include files to the search path at compile time.

Adds the library directory search path at link time

Enable the compiler to generate multi-threaded code based on the OpenMP* directives (same as -fopenmp)

Enables OpenMP* offloading compilation for target pragmas. This option only applies to Intel(R) MIC Architecture and Intel(R) Graphics Technology. Enabled by default with -qopenmp. Use -qno-openmp-offload to disable.
Specify kind to specify the default device for target pragmas
host - allow target code to run on host system while still doing the outlining for offload
mic - specify Intel(R) MIC Architecture
gfx - specify Intel(R) Graphics Technology

Enable -O3 -no-prec-div -fp-model fast=2 optimizations.

Optimize for maximum speed and enable more aggressive optimizations that may not improve performance on some programs.

Improve precision of FP divides (some speed impact).

-fp-model
enable floating point model variation
[no-]except - enable/disable floating point exception semantics
fast[=1|2] - enables more aggressive floating point optimizations
precise - allows value-safe optimizations
source - enables intermediates in source precision
sets -assume protect_parens for Fortran
strict - enables -fp-model precise -fp-model except, disables
contractions and enables pragma stdc fenv_access
consistent - enables consistent, reproducible results for
different optimization levels or between different
processors of the same architecture
double - rounds intermediates in 53-bit (double) precision
extended - rounds intermediates in 64-bit (extended) precision

May generate Intel(R) Advanced Vector Extensions 512 (Intel(R) AVX-512) Foundation instructions, Intel(R) AVX-512 Conflict Detection instructions, Intel(R) AVX-512 Doubleword and Quadword instructions, Intel(R) AVX-512 Byte and Word instructions and Intel(R) AVX-512 Vector Length Extensions for Intel(R) processors, and the instructions enabled with CORE-AVX2.

-qopt-zmm-usage=
Specifies the level of zmm registers usage. You can specify one of the following:
low - Tells the compiler that the compiled program is unlikely to benefit from zmm registers usage. It specifies that the compiler should avoid using zmm registers unless it can prove the gain from their usage.
high - Tells the compiler to generate zmm code without restrictions

-fimf-precision=value[:funclist]
defines the accuracy (precision) for math library functions
value - defined as one of the following values
high - equivalent to max-error = 0.6
medium - equivalent to max-error = 4 (DEFAULT)
low - equivalent to accuracy-bits = 11 (single precision); accuracy-bits = 26 (double precision)
funclist - optional comma separated list of one or more math library functions to which the attribute should be applied

Determine if certain square root optimizations are enabled.

Determine if certain square root optimizations are enabled.

Specifies whether streaming stores are generated:

always - enables generation of streaming stores under the assumption that the application is memory bound

auto - compiler decides when streaming stores are used (DEFAULT)

never - disables generation of streaming stores

Enable single-file IP optimization within files.

Enable multi-file IP optimization between files.

Enable levels of prefetch insertion, where 0 disables. n may be 0 through 5 inclusive. Default is 2.

Enable/disable(DEFAULT) use of ANSI aliasing rules in optimizations; user asserts that the program adheres to these rules.

Specifies preferred 512b vector width for auto-vectorization. Defaults to 'none' which allows target specific decisions.

Enable the compiler to generate multi-threaded code based on the OpenMP* directives. Similar behavior was granted by -qopenmp in previous versions.

Enable LTO (Link Time Optimization) in 'full' mode.

Allow aggressive, lossy floating-point optimizations.

Turn on loop unroller.

FPORTABILITY flag

No Fortran main method exists, use C equivalent instead.

Enables the use of nested SIMD statements for OpenMP.

Invoke the Intel C compiler.

Invoke the Intel C++ compiler.

Invoke the Intel Fortran compiler.

Invoke the Intel oneAPI C compiler.

Invoke the Intel oneAPI C++ compiler.

Invoke the Intel oneAPI Fortran compiler.

Link using FFTW 3.3.10 library for Linux. Description from FFTW:

FFTW lib compiled with -xCORE-AVX512 -qopt-zmm-usage=high -Ofast -fp-model fast=2

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions, of arbitrary input size, and of both real and complex data (as well as of even/odd data, i.e. the discrete cosine/sine transforms or DCT/DST).