CPU2017 Flag Description
Cisco Systems Cisco UCS B200 M5 (Intel Xeon Platinum 8170, 2.10 GHz)

Base Optimization Flags

C benchmarks

• • -xCORE-AVX512
  • 
  • COPTIMIZE

  • Code is optimized for Intel(R) processors with support for CORE-AVX512 instructions. The resulting code may contain unconditional use of features that are not supported on other processors. This option also enables new optimizations in addition to Intel processor-specific optimizations including advanced data layout and code restructuring optimizations to improve memory accesses for Intel processors.

    Do not use this option if you are executing a program on a processor that is not an Intel processor. If you use this option on a non-compatible processor to compile the main program (in Fortran) or the function main() in C/C++, the program will display a fatal run-time error if they are executed on unsupported processors.

• • -ipo
  • 
  • COPTIMIZE

  • Multi-file ip optimizations that includes:
    - inline function expansion
    - interprocedural constant propogation
    - dead code elimination
    - propagation of function characteristics
    - passing arguments in registers
    - loop-invariant code motion

• • -O3
  • 
  • COPTIMIZE

  • Enables O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enables optimizations for maximum speed, such as:

    • Loop unrolling, including instruction scheduling
    • Code replication to eliminate branches
    • Padding the size of certain power-of-two arrays to allow more efficient cache use.
    
    

    On IA-32 and Intel EM64T processors, when O3 is used with options -ax or -x (Linux) or with options /Qax or /Qx (Windows), the compiler performs more aggressive data dependency analysis than for O2, which may result in longer compilation times. The O3 optimizations may not cause higher performance unless loop and memory access transformations take place. The optimizations may slow down code in some cases compared to O2 optimizations. The O3 option is recommended for applications that have loops that heavily use floating-point calculations and process large data sets.

  • Includes:
    • -O2
      • -O1
        • -funroll-loops
        • -fno-builtin
        • -mno-ieee-fp
        • -fomit-framepointer
        • -ffunction-sections
        • -ftz
• • -no-prec-div
  • 
  • COPTIMIZE
  • (disable/enable[default] -prec-div)

    -no-prec-div enables optimizations that give slightly less precise results than full IEEE division.

    When you specify -no-prec-div along with some optimizations, such as -xN and -xB (Linux) or /QxN and /QxB (Windows), the compiler may change floating-point division computations into multiplication by the reciprocal of the denominator. For example, A/B is computed as A * (1/B) to improve the speed of the computation.

    However, sometimes the value produced by this transformation is not as accurate as full IEEE division. When it is important to have fully precise IEEE division, do not use -no-prec-div. This will enable the default -prec-div and the result will be more accurate, with some loss of performance.

• • -qopt-prefetch
  • 
  • COPTIMIZE

  • Enable/disable(DEFAULT) the compiler to generate prefetch instructions to prefetch data.

• • -ffinite-math-only
  • 
  • COPTIMIZE

  • Allow optimizations for floating point arithmetic that assume arguments and results are not NaNs or Infinities

• • -qopt-mem-layout-trans=3
  • 
  • COPTIMIZE

  • Controls the level of memory layout transformations performed by the compiler. This option can improve cache reuse and cache locality.

    • 0: Disables memory layout transformations. This is the same as specifying -qno-opt-mem-layout-trans
    • 1: Enables basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enables more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Compiler is more aggressive in using memory layout transformations. You should only use this setting if your system has more than 4GB of physical memory per core.
• • -qopenmp
  • 
  • Yes
  • COPTIMIZE

• • -DSPEC_OPENMP
  • 
  • COPTIMIZE

  • Definition of this macro indicates that compilation for parallel operation is enabled, and that any OpenMP directives or pragmas will be visible to the compiler. The behavior of this macro is overridden if -DSPEC_SUPPRESS_OPENMP also appears in the list of compilation flags.

Fortran benchmarks

• • -DSPEC_OPENMP
  • 
  • FOPTIMIZE

  • Definition of this macro indicates that compilation for parallel operation is enabled, and that any OpenMP directives or pragmas will be visible to the compiler. The behavior of this macro is overridden if -DSPEC_SUPPRESS_OPENMP also appears in the list of compilation flags.

• • -xCORE-AVX512
  • 
  • FOPTIMIZE

  • Code is optimized for Intel(R) processors with support for CORE-AVX512 instructions. The resulting code may contain unconditional use of features that are not supported on other processors. This option also enables new optimizations in addition to Intel processor-specific optimizations including advanced data layout and code restructuring optimizations to improve memory accesses for Intel processors.

    Do not use this option if you are executing a program on a processor that is not an Intel processor. If you use this option on a non-compatible processor to compile the main program (in Fortran) or the function main() in C/C++, the program will display a fatal run-time error if they are executed on unsupported processors.

• • -ipo
  • 
  • FOPTIMIZE

  • Multi-file ip optimizations that includes:
    - inline function expansion
    - interprocedural constant propogation
    - dead code elimination
    - propagation of function characteristics
    - passing arguments in registers
    - loop-invariant code motion

• • -O3
  • 
  • FOPTIMIZE

  • Enables O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enables optimizations for maximum speed, such as:

    • Loop unrolling, including instruction scheduling
    • Code replication to eliminate branches
    • Padding the size of certain power-of-two arrays to allow more efficient cache use.
    
    

    On IA-32 and Intel EM64T processors, when O3 is used with options -ax or -x (Linux) or with options /Qax or /Qx (Windows), the compiler performs more aggressive data dependency analysis than for O2, which may result in longer compilation times. The O3 optimizations may not cause higher performance unless loop and memory access transformations take place. The optimizations may slow down code in some cases compared to O2 optimizations. The O3 option is recommended for applications that have loops that heavily use floating-point calculations and process large data sets.

  • Includes:
    • -O2
      • -O1
        • -funroll-loops
        • -fno-builtin
        • -mno-ieee-fp
        • -fomit-framepointer
        • -ffunction-sections
        • -ftz
• • -no-prec-div
  • 
  • FOPTIMIZE
  • (disable/enable[default] -prec-div)

    -no-prec-div enables optimizations that give slightly less precise results than full IEEE division.

    When you specify -no-prec-div along with some optimizations, such as -xN and -xB (Linux) or /QxN and /QxB (Windows), the compiler may change floating-point division computations into multiplication by the reciprocal of the denominator. For example, A/B is computed as A * (1/B) to improve the speed of the computation.

    However, sometimes the value produced by this transformation is not as accurate as full IEEE division. When it is important to have fully precise IEEE division, do not use -no-prec-div. This will enable the default -prec-div and the result will be more accurate, with some loss of performance.

• • -qopt-prefetch
  • 
  • FOPTIMIZE

  • Enable/disable(DEFAULT) the compiler to generate prefetch instructions to prefetch data.

• • -ffinite-math-only
  • 
  • FOPTIMIZE

  • Allow optimizations for floating point arithmetic that assume arguments and results are not NaNs or Infinities

• • -qopt-mem-layout-trans=3
  • 
  • FOPTIMIZE

  • Controls the level of memory layout transformations performed by the compiler. This option can improve cache reuse and cache locality.

    • 0: Disables memory layout transformations. This is the same as specifying -qno-opt-mem-layout-trans
    • 1: Enables basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enables more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Compiler is more aggressive in using memory layout transformations. You should only use this setting if your system has more than 4GB of physical memory per core.
• • -qopenmp
  • 
  • Yes
  • FOPTIMIZE

• • -nostandard-realloc-lhs
  • 
  • EXTRA_FOPTIMIZE

  • Option standard-realloc-lhs (the default), tells the compiler that when the left-hand side of an assignment is an allocatable object, it should be reallocated to the shape of the right-hand side of the assignment before the assignment occurs. This is the current Fortran Standard definition. This feature may cause extra overhead at run time. This option has the same effect as option assume realloc_lhs.

    If you specify nostandard-realloc-lhs, the compiler uses the old Fortran 2003 rules when interpreting assignment statements. The left-hand side is assumed to be allocated with the correct shape to hold the right-hand side. If it is not, incorrect behavior will occur. This option has the same effect as option assume norealloc_lhs.

• • -align array32byte
  • 
  • EXTRA_FOPTIMIZE

  • The align toggle changes how data elements are aligned. Variables and arrays are analyzed and memory layout can be altered. Specifying array32byte will look for opportunities to transform and reailgn arrays to 32byte boundaries.

Benchmarks using both Fortran and C

• • -xCORE-AVX512
  • 
  • COPTIMIZE, FOPTIMIZE

  • Code is optimized for Intel(R) processors with support for CORE-AVX512 instructions. The resulting code may contain unconditional use of features that are not supported on other processors. This option also enables new optimizations in addition to Intel processor-specific optimizations including advanced data layout and code restructuring optimizations to improve memory accesses for Intel processors.

    Do not use this option if you are executing a program on a processor that is not an Intel processor. If you use this option on a non-compatible processor to compile the main program (in Fortran) or the function main() in C/C++, the program will display a fatal run-time error if they are executed on unsupported processors.

• • -ipo
  • 
  • COPTIMIZE, FOPTIMIZE

  • Multi-file ip optimizations that includes:
    - inline function expansion
    - interprocedural constant propogation
    - dead code elimination
    - propagation of function characteristics
    - passing arguments in registers
    - loop-invariant code motion

• • -O3
  • 
  • COPTIMIZE, FOPTIMIZE

  • Enables O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enables optimizations for maximum speed, such as:

    • Loop unrolling, including instruction scheduling
    • Code replication to eliminate branches
    • Padding the size of certain power-of-two arrays to allow more efficient cache use.
    
    

    On IA-32 and Intel EM64T processors, when O3 is used with options -ax or -x (Linux) or with options /Qax or /Qx (Windows), the compiler performs more aggressive data dependency analysis than for O2, which may result in longer compilation times. The O3 optimizations may not cause higher performance unless loop and memory access transformations take place. The optimizations may slow down code in some cases compared to O2 optimizations. The O3 option is recommended for applications that have loops that heavily use floating-point calculations and process large data sets.

  • Includes:
    • -O2
      • -O1
        • -funroll-loops
        • -fno-builtin
        • -mno-ieee-fp
        • -fomit-framepointer
        • -ffunction-sections
        • -ftz
• • -no-prec-div
  • 
  • COPTIMIZE, FOPTIMIZE
  • (disable/enable[default] -prec-div)

    -no-prec-div enables optimizations that give slightly less precise results than full IEEE division.

    When you specify -no-prec-div along with some optimizations, such as -xN and -xB (Linux) or /QxN and /QxB (Windows), the compiler may change floating-point division computations into multiplication by the reciprocal of the denominator. For example, A/B is computed as A * (1/B) to improve the speed of the computation.

    However, sometimes the value produced by this transformation is not as accurate as full IEEE division. When it is important to have fully precise IEEE division, do not use -no-prec-div. This will enable the default -prec-div and the result will be more accurate, with some loss of performance.

• • -qopt-prefetch
  • 
  • COPTIMIZE, FOPTIMIZE

  • Enable/disable(DEFAULT) the compiler to generate prefetch instructions to prefetch data.

• • -ffinite-math-only
  • 
  • COPTIMIZE, FOPTIMIZE

  • Allow optimizations for floating point arithmetic that assume arguments and results are not NaNs or Infinities

• • -qopt-mem-layout-trans=3
  • 
  • COPTIMIZE, FOPTIMIZE

  • Controls the level of memory layout transformations performed by the compiler. This option can improve cache reuse and cache locality.

    • 0: Disables memory layout transformations. This is the same as specifying -qno-opt-mem-layout-trans
    • 1: Enables basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enables more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Compiler is more aggressive in using memory layout transformations. You should only use this setting if your system has more than 4GB of physical memory per core.
• • -qopenmp
  • 
  • Yes
  • COPTIMIZE, FOPTIMIZE

• • -DSPEC_OPENMP
  • 
  • COPTIMIZE, FOPTIMIZE

  • Definition of this macro indicates that compilation for parallel operation is enabled, and that any OpenMP directives or pragmas will be visible to the compiler. The behavior of this macro is overridden if -DSPEC_SUPPRESS_OPENMP also appears in the list of compilation flags.

• • -nostandard-realloc-lhs
  • 
  • EXTRA_FOPTIMIZE

  • Option standard-realloc-lhs (the default), tells the compiler that when the left-hand side of an assignment is an allocatable object, it should be reallocated to the shape of the right-hand side of the assignment before the assignment occurs. This is the current Fortran Standard definition. This feature may cause extra overhead at run time. This option has the same effect as option assume realloc_lhs.

    If you specify nostandard-realloc-lhs, the compiler uses the old Fortran 2003 rules when interpreting assignment statements. The left-hand side is assumed to be allocated with the correct shape to hold the right-hand side. If it is not, incorrect behavior will occur. This option has the same effect as option assume norealloc_lhs.

• • -align array32byte
  • 
  • EXTRA_FOPTIMIZE

  • The align toggle changes how data elements are aligned. Variables and arrays are analyzed and memory layout can be altered. Specifying array32byte will look for opportunities to transform and reailgn arrays to 32byte boundaries.

Benchmarks using Fortran, C, and C++

• • -xCORE-AVX512
  • 
  • COPTIMIZE, CXXOPTIMIZE, FOPTIMIZE

  • Code is optimized for Intel(R) processors with support for CORE-AVX512 instructions. The resulting code may contain unconditional use of features that are not supported on other processors. This option also enables new optimizations in addition to Intel processor-specific optimizations including advanced data layout and code restructuring optimizations to improve memory accesses for Intel processors.

    Do not use this option if you are executing a program on a processor that is not an Intel processor. If you use this option on a non-compatible processor to compile the main program (in Fortran) or the function main() in C/C++, the program will display a fatal run-time error if they are executed on unsupported processors.

• • -ipo
  • 
  • COPTIMIZE, CXXOPTIMIZE, FOPTIMIZE

  • Multi-file ip optimizations that includes:
    - inline function expansion
    - interprocedural constant propogation
    - dead code elimination
    - propagation of function characteristics
    - passing arguments in registers
    - loop-invariant code motion

• • -O3
  • 
  • COPTIMIZE, CXXOPTIMIZE, FOPTIMIZE

  • Enables O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enables optimizations for maximum speed, such as:

    • Loop unrolling, including instruction scheduling
    • Code replication to eliminate branches
    • Padding the size of certain power-of-two arrays to allow more efficient cache use.
    
    

    On IA-32 and Intel EM64T processors, when O3 is used with options -ax or -x (Linux) or with options /Qax or /Qx (Windows), the compiler performs more aggressive data dependency analysis than for O2, which may result in longer compilation times. The O3 optimizations may not cause higher performance unless loop and memory access transformations take place. The optimizations may slow down code in some cases compared to O2 optimizations. The O3 option is recommended for applications that have loops that heavily use floating-point calculations and process large data sets.

  • Includes:
    • -O2
      • -O1
        • -funroll-loops
        • -fno-builtin
        • -mno-ieee-fp
        • -fomit-framepointer
        • -ffunction-sections
        • -ftz
• • -no-prec-div
  • 
  • COPTIMIZE, CXXOPTIMIZE, FOPTIMIZE
  • (disable/enable[default] -prec-div)

    -no-prec-div enables optimizations that give slightly less precise results than full IEEE division.

    When you specify -no-prec-div along with some optimizations, such as -xN and -xB (Linux) or /QxN and /QxB (Windows), the compiler may change floating-point division computations into multiplication by the reciprocal of the denominator. For example, A/B is computed as A * (1/B) to improve the speed of the computation.

    However, sometimes the value produced by this transformation is not as accurate as full IEEE division. When it is important to have fully precise IEEE division, do not use -no-prec-div. This will enable the default -prec-div and the result will be more accurate, with some loss of performance.

• • -qopt-prefetch
  • 
  • COPTIMIZE, CXXOPTIMIZE, FOPTIMIZE

  • Enable/disable(DEFAULT) the compiler to generate prefetch instructions to prefetch data.

• • -ffinite-math-only
  • 
  • COPTIMIZE, CXXOPTIMIZE, FOPTIMIZE

  • Allow optimizations for floating point arithmetic that assume arguments and results are not NaNs or Infinities

• • -qopt-mem-layout-trans=3
  • 
  • COPTIMIZE, CXXOPTIMIZE, FOPTIMIZE

  • Controls the level of memory layout transformations performed by the compiler. This option can improve cache reuse and cache locality.

    • 0: Disables memory layout transformations. This is the same as specifying -qno-opt-mem-layout-trans
    • 1: Enables basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enables more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Compiler is more aggressive in using memory layout transformations. You should only use this setting if your system has more than 4GB of physical memory per core.
• • -qopenmp
  • 
  • Yes
  • COPTIMIZE, CXXOPTIMIZE, FOPTIMIZE

• • -DSPEC_OPENMP
  • 
  • COPTIMIZE, CXXOPTIMIZE, FOPTIMIZE

  • Definition of this macro indicates that compilation for parallel operation is enabled, and that any OpenMP directives or pragmas will be visible to the compiler. The behavior of this macro is overridden if -DSPEC_SUPPRESS_OPENMP also appears in the list of compilation flags.

• • -nostandard-realloc-lhs
  • 
  • EXTRA_FOPTIMIZE

  • Option standard-realloc-lhs (the default), tells the compiler that when the left-hand side of an assignment is an allocatable object, it should be reallocated to the shape of the right-hand side of the assignment before the assignment occurs. This is the current Fortran Standard definition. This feature may cause extra overhead at run time. This option has the same effect as option assume realloc_lhs.

    If you specify nostandard-realloc-lhs, the compiler uses the old Fortran 2003 rules when interpreting assignment statements. The left-hand side is assumed to be allocated with the correct shape to hold the right-hand side. If it is not, incorrect behavior will occur. This option has the same effect as option assume norealloc_lhs.

• • -align array32byte
  • 
  • EXTRA_FOPTIMIZE

  • The align toggle changes how data elements are aligned. Variables and arrays are analyzed and memory layout can be altered. Specifying array32byte will look for opportunities to transform and reailgn arrays to 32byte boundaries.

Peak Optimization Flags

C benchmarks

619.lbm_s

• • -prof-gen
  • 
  • PASS1_CFLAGS, PASS1_LDFLAGS

  • Instrument program for profiling for the first phase of two-phase profile guided otimization. This instrumentation gathers information about a program's execution paths and data values but does not gather information from hardware performance counters. The profile instrumentation also gathers data for optimizations which are unique to profile-feedback optimization.

    -profgen:threadsafe option collects profile guided optimization data with guards for threaded applications.

• • -prof-use
  • 
  • PASS2_CFLAGS, PASS2_LDFLAGS

  • Instructs the compiler to produce a profile-optimized executable and merges available dynamic information (.dyn) files into a pgopti.dpi file. If you perform multiple executions of the instrumented program, -prof-use merges the dynamic information files again and overwrites the previous pgopti.dpi file.
    Without any other options, the current directory is searched for .dyn files

• • -O2
  • 
  • PASS1_COPTIMIZE

  • Enables optimizations for speed. This is the generally recommended optimization level. This option also enables:
    - Inlining of intrinsics
    - Intra-file interprocedural optimizations, which include:
    - inlining
    - constant propagation
    - forward substitution
    - routine attribute propagation
    - variable address-taken analysis
    - dead static function elimination
    - removal of unreferenced variables
    - The following capabilities for performance gain:
    - constant propagation
    - copy propagation
    - dead-code elimination
    - global register allocation
    - global instruction scheduling and control speculation
    - loop unrolling
    - optimized code selection
    - partial redundancy elimination
    - strength reduction/induction variable simplification
    - variable renaming
    - exception handling optimizations
    - tail recursions
    - peephole optimizations
    - structure assignment lowering and optimizations
    - dead store elimination
    

  • Includes:
    • -O1
      • -funroll-loops
      • -fno-builtin
      • -mno-ieee-fp
      • -fomit-framepointer
      • -ffunction-sections
      • -ftz
• • -xCORE-AVX512
  • 
  • PASS2_COPTIMIZE

  • Code is optimized for Intel(R) processors with support for CORE-AVX512 instructions. The resulting code may contain unconditional use of features that are not supported on other processors. This option also enables new optimizations in addition to Intel processor-specific optimizations including advanced data layout and code restructuring optimizations to improve memory accesses for Intel processors.

    Do not use this option if you are executing a program on a processor that is not an Intel processor. If you use this option on a non-compatible processor to compile the main program (in Fortran) or the function main() in C/C++, the program will display a fatal run-time error if they are executed on unsupported processors.

• • -qopt-prefetch
  • 
  • PASS1_COPTIMIZE, PASS2_COPTIMIZE

  • Enable/disable(DEFAULT) the compiler to generate prefetch instructions to prefetch data.

• • -ipo
  • 
  • PASS2_COPTIMIZE

  • Multi-file ip optimizations that includes:
    - inline function expansion
    - interprocedural constant propogation
    - dead code elimination
    - propagation of function characteristics
    - passing arguments in registers
    - loop-invariant code motion

• • -O3
  • 
  • PASS2_COPTIMIZE

  • Enables O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enables optimizations for maximum speed, such as:

    • Loop unrolling, including instruction scheduling
    • Code replication to eliminate branches
    • Padding the size of certain power-of-two arrays to allow more efficient cache use.
    
    

    On IA-32 and Intel EM64T processors, when O3 is used with options -ax or -x (Linux) or with options /Qax or /Qx (Windows), the compiler performs more aggressive data dependency analysis than for O2, which may result in longer compilation times. The O3 optimizations may not cause higher performance unless loop and memory access transformations take place. The optimizations may slow down code in some cases compared to O2 optimizations. The O3 option is recommended for applications that have loops that heavily use floating-point calculations and process large data sets.

  • Includes:
    • -O2
      • -O1
        • -funroll-loops
        • -fno-builtin
        • -mno-ieee-fp
        • -fomit-framepointer
        • -ffunction-sections
        • -ftz
• • -ffinite-math-only
  • 
  • PASS1_COPTIMIZE, PASS2_COPTIMIZE

  • Allow optimizations for floating point arithmetic that assume arguments and results are not NaNs or Infinities

• • -no-prec-div
  • 
  • PASS2_COPTIMIZE
  • (disable/enable[default] -prec-div)

    -no-prec-div enables optimizations that give slightly less precise results than full IEEE division.

    When you specify -no-prec-div along with some optimizations, such as -xN and -xB (Linux) or /QxN and /QxB (Windows), the compiler may change floating-point division computations into multiplication by the reciprocal of the denominator. For example, A/B is computed as A * (1/B) to improve the speed of the computation.

    However, sometimes the value produced by this transformation is not as accurate as full IEEE division. When it is important to have fully precise IEEE division, do not use -no-prec-div. This will enable the default -prec-div and the result will be more accurate, with some loss of performance.

• • -qopt-mem-layout-trans=3
  • 
  • PASS1_COPTIMIZE, PASS2_COPTIMIZE

  • Controls the level of memory layout transformations performed by the compiler. This option can improve cache reuse and cache locality.

    • 0: Disables memory layout transformations. This is the same as specifying -qno-opt-mem-layout-trans
    • 1: Enables basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enables more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Compiler is more aggressive in using memory layout transformations. You should only use this setting if your system has more than 4GB of physical memory per core.
• • -DSPEC_SUPPRESS_OPENMP
  • 
  • PASS1_COPTIMIZE

  • Definition of this macro indicates that compilation for parallel operation is disabled, and that any OpenMP directives or pragmas will not be visible to the compiler. This macro overrides the behavior of -DSPEC_OPENMP when they both appear in the list of compilation flags.

• • -qopenmp
  • 
  • Yes
  • PASS2_COPTIMIZE

• • -DSPEC_OPENMP
  • 
  • PASS2_COPTIMIZE

  • Definition of this macro indicates that compilation for parallel operation is enabled, and that any OpenMP directives or pragmas will be visible to the compiler. The behavior of this macro is overridden if -DSPEC_SUPPRESS_OPENMP also appears in the list of compilation flags.

638.imagick_s

• • -xCORE-AVX512
  • 
  • COPTIMIZE

  • Code is optimized for Intel(R) processors with support for CORE-AVX512 instructions. The resulting code may contain unconditional use of features that are not supported on other processors. This option also enables new optimizations in addition to Intel processor-specific optimizations including advanced data layout and code restructuring optimizations to improve memory accesses for Intel processors.

    Do not use this option if you are executing a program on a processor that is not an Intel processor. If you use this option on a non-compatible processor to compile the main program (in Fortran) or the function main() in C/C++, the program will display a fatal run-time error if they are executed on unsupported processors.

• • -ipo
  • 
  • COPTIMIZE

  • Multi-file ip optimizations that includes:
    - inline function expansion
    - interprocedural constant propogation
    - dead code elimination
    - propagation of function characteristics
    - passing arguments in registers
    - loop-invariant code motion

• • -O3
  • 
  • COPTIMIZE

  • Enables O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enables optimizations for maximum speed, such as:

    • Loop unrolling, including instruction scheduling
    • Code replication to eliminate branches
    • Padding the size of certain power-of-two arrays to allow more efficient cache use.
    
    

    On IA-32 and Intel EM64T processors, when O3 is used with options -ax or -x (Linux) or with options /Qax or /Qx (Windows), the compiler performs more aggressive data dependency analysis than for O2, which may result in longer compilation times. The O3 optimizations may not cause higher performance unless loop and memory access transformations take place. The optimizations may slow down code in some cases compared to O2 optimizations. The O3 option is recommended for applications that have loops that heavily use floating-point calculations and process large data sets.

  • Includes:
    • -O2
      • -O1
        • -funroll-loops
        • -fno-builtin
        • -mno-ieee-fp
        • -fomit-framepointer
        • -ffunction-sections
        • -ftz
• • -no-prec-div
  • 
  • COPTIMIZE
  • (disable/enable[default] -prec-div)

    -no-prec-div enables optimizations that give slightly less precise results than full IEEE division.

    When you specify -no-prec-div along with some optimizations, such as -xN and -xB (Linux) or /QxN and /QxB (Windows), the compiler may change floating-point division computations into multiplication by the reciprocal of the denominator. For example, A/B is computed as A * (1/B) to improve the speed of the computation.

    However, sometimes the value produced by this transformation is not as accurate as full IEEE division. When it is important to have fully precise IEEE division, do not use -no-prec-div. This will enable the default -prec-div and the result will be more accurate, with some loss of performance.

• • -qopt-prefetch
  • 
  • COPTIMIZE

  • Enable/disable(DEFAULT) the compiler to generate prefetch instructions to prefetch data.

• • -ffinite-math-only
  • 
  • COPTIMIZE

  • Allow optimizations for floating point arithmetic that assume arguments and results are not NaNs or Infinities

• • -qopt-mem-layout-trans=3
  • 
  • COPTIMIZE

  • Controls the level of memory layout transformations performed by the compiler. This option can improve cache reuse and cache locality.

    • 0: Disables memory layout transformations. This is the same as specifying -qno-opt-mem-layout-trans
    • 1: Enables basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enables more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Compiler is more aggressive in using memory layout transformations. You should only use this setting if your system has more than 4GB of physical memory per core.
• • -qopenmp
  • 
  • Yes
  • COPTIMIZE

• • -DSPEC_OPENMP
  • 
  • COPTIMIZE

  • Definition of this macro indicates that compilation for parallel operation is enabled, and that any OpenMP directives or pragmas will be visible to the compiler. The behavior of this macro is overridden if -DSPEC_SUPPRESS_OPENMP also appears in the list of compilation flags.

644.nab_s

• • Same as 638.imagick_s

Fortran benchmarks

• • -prof-gen
  • 
  • PASS1_FFLAGS, PASS1_LDFLAGS

  • Instrument program for profiling for the first phase of two-phase profile guided otimization. This instrumentation gathers information about a program's execution paths and data values but does not gather information from hardware performance counters. The profile instrumentation also gathers data for optimizations which are unique to profile-feedback optimization.

    -profgen:threadsafe option collects profile guided optimization data with guards for threaded applications.

• • -prof-use
  • 
  • PASS2_FFLAGS, PASS2_LDFLAGS

  • Instructs the compiler to produce a profile-optimized executable and merges available dynamic information (.dyn) files into a pgopti.dpi file. If you perform multiple executions of the instrumented program, -prof-use merges the dynamic information files again and overwrites the previous pgopti.dpi file.
    Without any other options, the current directory is searched for .dyn files

• • -DSPEC_SUPPRESS_OPENMP
  • 
  • PASS1_FOPTIMIZE

  • Definition of this macro indicates that compilation for parallel operation is disabled, and that any OpenMP directives or pragmas will not be visible to the compiler. This macro overrides the behavior of -DSPEC_OPENMP when they both appear in the list of compilation flags.

• • -DSPEC_OPENMP
  • 
  • PASS2_FOPTIMIZE

  • Definition of this macro indicates that compilation for parallel operation is enabled, and that any OpenMP directives or pragmas will be visible to the compiler. The behavior of this macro is overridden if -DSPEC_SUPPRESS_OPENMP also appears in the list of compilation flags.

• • -O2
  • 
  • PASS1_FOPTIMIZE

  • Enables optimizations for speed. This is the generally recommended optimization level. This option also enables:
    - Inlining of intrinsics
    - Intra-file interprocedural optimizations, which include:
    - inlining
    - constant propagation
    - forward substitution
    - routine attribute propagation
    - variable address-taken analysis
    - dead static function elimination
    - removal of unreferenced variables
    - The following capabilities for performance gain:
    - constant propagation
    - copy propagation
    - dead-code elimination
    - global register allocation
    - global instruction scheduling and control speculation
    - loop unrolling
    - optimized code selection
    - partial redundancy elimination
    - strength reduction/induction variable simplification
    - variable renaming
    - exception handling optimizations
    - tail recursions
    - peephole optimizations
    - structure assignment lowering and optimizations
    - dead store elimination
    

  • Includes:
    • -O1
      • -funroll-loops
      • -fno-builtin
      • -mno-ieee-fp
      • -fomit-framepointer
      • -ffunction-sections
      • -ftz
• • -xCORE-AVX512
  • 
  • PASS2_FOPTIMIZE

  • Code is optimized for Intel(R) processors with support for CORE-AVX512 instructions. The resulting code may contain unconditional use of features that are not supported on other processors. This option also enables new optimizations in addition to Intel processor-specific optimizations including advanced data layout and code restructuring optimizations to improve memory accesses for Intel processors.

    Do not use this option if you are executing a program on a processor that is not an Intel processor. If you use this option on a non-compatible processor to compile the main program (in Fortran) or the function main() in C/C++, the program will display a fatal run-time error if they are executed on unsupported processors.

• • -qopt-prefetch
  • 
  • PASS1_FOPTIMIZE, PASS2_FOPTIMIZE

  • Enable/disable(DEFAULT) the compiler to generate prefetch instructions to prefetch data.

• • -ipo
  • 
  • PASS2_FOPTIMIZE

  • Multi-file ip optimizations that includes:
    - inline function expansion
    - interprocedural constant propogation
    - dead code elimination
    - propagation of function characteristics
    - passing arguments in registers
    - loop-invariant code motion

• • -O3
  • 
  • PASS2_FOPTIMIZE

  • Enables O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enables optimizations for maximum speed, such as:

    • Loop unrolling, including instruction scheduling
    • Code replication to eliminate branches
    • Padding the size of certain power-of-two arrays to allow more efficient cache use.
    
    

    On IA-32 and Intel EM64T processors, when O3 is used with options -ax or -x (Linux) or with options /Qax or /Qx (Windows), the compiler performs more aggressive data dependency analysis than for O2, which may result in longer compilation times. The O3 optimizations may not cause higher performance unless loop and memory access transformations take place. The optimizations may slow down code in some cases compared to O2 optimizations. The O3 option is recommended for applications that have loops that heavily use floating-point calculations and process large data sets.

  • Includes:
    • -O2
      • -O1
        • -funroll-loops
        • -fno-builtin
        • -mno-ieee-fp
        • -fomit-framepointer
        • -ffunction-sections
        • -ftz
• • -ffinite-math-only
  • 
  • PASS1_FOPTIMIZE, PASS2_FOPTIMIZE

  • Allow optimizations for floating point arithmetic that assume arguments and results are not NaNs or Infinities

• • -no-prec-div
  • 
  • PASS2_FOPTIMIZE
  • (disable/enable[default] -prec-div)

    -no-prec-div enables optimizations that give slightly less precise results than full IEEE division.

    When you specify -no-prec-div along with some optimizations, such as -xN and -xB (Linux) or /QxN and /QxB (Windows), the compiler may change floating-point division computations into multiplication by the reciprocal of the denominator. For example, A/B is computed as A * (1/B) to improve the speed of the computation.

    However, sometimes the value produced by this transformation is not as accurate as full IEEE division. When it is important to have fully precise IEEE division, do not use -no-prec-div. This will enable the default -prec-div and the result will be more accurate, with some loss of performance.

• • -qopt-mem-layout-trans=3
  • 
  • PASS1_FOPTIMIZE, PASS2_FOPTIMIZE

  • Controls the level of memory layout transformations performed by the compiler. This option can improve cache reuse and cache locality.

    • 0: Disables memory layout transformations. This is the same as specifying -qno-opt-mem-layout-trans
    • 1: Enables basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enables more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Compiler is more aggressive in using memory layout transformations. You should only use this setting if your system has more than 4GB of physical memory per core.
• • -qopenmp
  • 
  • Yes
  • PASS2_FOPTIMIZE

• • -nostandard-realloc-lhs
  • 
  • EXTRA_FOPTIMIZE

  • Option standard-realloc-lhs (the default), tells the compiler that when the left-hand side of an assignment is an allocatable object, it should be reallocated to the shape of the right-hand side of the assignment before the assignment occurs. This is the current Fortran Standard definition. This feature may cause extra overhead at run time. This option has the same effect as option assume realloc_lhs.

    If you specify nostandard-realloc-lhs, the compiler uses the old Fortran 2003 rules when interpreting assignment statements. The left-hand side is assumed to be allocated with the correct shape to hold the right-hand side. If it is not, incorrect behavior will occur. This option has the same effect as option assume norealloc_lhs.

• • -align array32byte
  • 
  • EXTRA_FOPTIMIZE

  • The align toggle changes how data elements are aligned. Variables and arrays are analyzed and memory layout can be altered. Specifying array32byte will look for opportunities to transform and reailgn arrays to 32byte boundaries.

Benchmarks using both Fortran and C

621.wrf_s

• • -prof-gen
  • 
  • PASS1_CFLAGS, PASS1_FFLAGS, PASS1_LDFLAGS

  • Instrument program for profiling for the first phase of two-phase profile guided otimization. This instrumentation gathers information about a program's execution paths and data values but does not gather information from hardware performance counters. The profile instrumentation also gathers data for optimizations which are unique to profile-feedback optimization.

    -profgen:threadsafe option collects profile guided optimization data with guards for threaded applications.

• • -prof-use
  • 
  • PASS2_CFLAGS, PASS2_FFLAGS, PASS2_LDFLAGS

  • Instructs the compiler to produce a profile-optimized executable and merges available dynamic information (.dyn) files into a pgopti.dpi file. If you perform multiple executions of the instrumented program, -prof-use merges the dynamic information files again and overwrites the previous pgopti.dpi file.
    Without any other options, the current directory is searched for .dyn files

• • -O2
  • 
  • PASS1_COPTIMIZE, PASS1_FOPTIMIZE

  • Enables optimizations for speed. This is the generally recommended optimization level. This option also enables:
    - Inlining of intrinsics
    - Intra-file interprocedural optimizations, which include:
    - inlining
    - constant propagation
    - forward substitution
    - routine attribute propagation
    - variable address-taken analysis
    - dead static function elimination
    - removal of unreferenced variables
    - The following capabilities for performance gain:
    - constant propagation
    - copy propagation
    - dead-code elimination
    - global register allocation
    - global instruction scheduling and control speculation
    - loop unrolling
    - optimized code selection
    - partial redundancy elimination
    - strength reduction/induction variable simplification
    - variable renaming
    - exception handling optimizations
    - tail recursions
    - peephole optimizations
    - structure assignment lowering and optimizations
    - dead store elimination
    

  • Includes:
    • -O1
      • -funroll-loops
      • -fno-builtin
      • -mno-ieee-fp
      • -fomit-framepointer
      • -ffunction-sections
      • -ftz
• • -xCORE-AVX512
  • 
  • PASS2_COPTIMIZE, PASS2_FOPTIMIZE

  • Code is optimized for Intel(R) processors with support for CORE-AVX512 instructions. The resulting code may contain unconditional use of features that are not supported on other processors. This option also enables new optimizations in addition to Intel processor-specific optimizations including advanced data layout and code restructuring optimizations to improve memory accesses for Intel processors.

    Do not use this option if you are executing a program on a processor that is not an Intel processor. If you use this option on a non-compatible processor to compile the main program (in Fortran) or the function main() in C/C++, the program will display a fatal run-time error if they are executed on unsupported processors.

• • -qopt-prefetch
  • 
  • PASS1_COPTIMIZE, PASS1_FOPTIMIZE, PASS2_COPTIMIZE, PASS2_FOPTIMIZE

  • Enable/disable(DEFAULT) the compiler to generate prefetch instructions to prefetch data.

• • -ipo
  • 
  • PASS2_COPTIMIZE, PASS2_FOPTIMIZE

  • Multi-file ip optimizations that includes:
    - inline function expansion
    - interprocedural constant propogation
    - dead code elimination
    - propagation of function characteristics
    - passing arguments in registers
    - loop-invariant code motion

• • -O3
  • 
  • PASS2_COPTIMIZE, PASS2_FOPTIMIZE

  • Enables O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enables optimizations for maximum speed, such as:

    • Loop unrolling, including instruction scheduling
    • Code replication to eliminate branches
    • Padding the size of certain power-of-two arrays to allow more efficient cache use.
    
    

    On IA-32 and Intel EM64T processors, when O3 is used with options -ax or -x (Linux) or with options /Qax or /Qx (Windows), the compiler performs more aggressive data dependency analysis than for O2, which may result in longer compilation times. The O3 optimizations may not cause higher performance unless loop and memory access transformations take place. The optimizations may slow down code in some cases compared to O2 optimizations. The O3 option is recommended for applications that have loops that heavily use floating-point calculations and process large data sets.

  • Includes:
    • -O2
      • -O1
        • -funroll-loops
        • -fno-builtin
        • -mno-ieee-fp
        • -fomit-framepointer
        • -ffunction-sections
        • -ftz
• • -ffinite-math-only
  • 
  • PASS1_COPTIMIZE, PASS1_FOPTIMIZE, PASS2_COPTIMIZE, PASS2_FOPTIMIZE

  • Allow optimizations for floating point arithmetic that assume arguments and results are not NaNs or Infinities

• • -no-prec-div
  • 
  • PASS2_COPTIMIZE, PASS2_FOPTIMIZE
  • (disable/enable[default] -prec-div)

    -no-prec-div enables optimizations that give slightly less precise results than full IEEE division.

    When you specify -no-prec-div along with some optimizations, such as -xN and -xB (Linux) or /QxN and /QxB (Windows), the compiler may change floating-point division computations into multiplication by the reciprocal of the denominator. For example, A/B is computed as A * (1/B) to improve the speed of the computation.

    However, sometimes the value produced by this transformation is not as accurate as full IEEE division. When it is important to have fully precise IEEE division, do not use -no-prec-div. This will enable the default -prec-div and the result will be more accurate, with some loss of performance.

• • -qopt-mem-layout-trans=3
  • 
  • PASS1_COPTIMIZE, PASS1_FOPTIMIZE, PASS2_COPTIMIZE, PASS2_FOPTIMIZE

  • Controls the level of memory layout transformations performed by the compiler. This option can improve cache reuse and cache locality.

    • 0: Disables memory layout transformations. This is the same as specifying -qno-opt-mem-layout-trans
    • 1: Enables basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enables more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Compiler is more aggressive in using memory layout transformations. You should only use this setting if your system has more than 4GB of physical memory per core.
• • -DSPEC_SUPPRESS_OPENMP
  • 
  • PASS1_COPTIMIZE, PASS1_FOPTIMIZE

  • Definition of this macro indicates that compilation for parallel operation is disabled, and that any OpenMP directives or pragmas will not be visible to the compiler. This macro overrides the behavior of -DSPEC_OPENMP when they both appear in the list of compilation flags.

• • -qopenmp
  • 
  • Yes
  • PASS2_COPTIMIZE, PASS2_FOPTIMIZE

• • -DSPEC_OPENMP
  • 
  • PASS2_COPTIMIZE, PASS2_FOPTIMIZE

  • Definition of this macro indicates that compilation for parallel operation is enabled, and that any OpenMP directives or pragmas will be visible to the compiler. The behavior of this macro is overridden if -DSPEC_SUPPRESS_OPENMP also appears in the list of compilation flags.

• • -nostandard-realloc-lhs
  • 
  • EXTRA_FOPTIMIZE

  • Option standard-realloc-lhs (the default), tells the compiler that when the left-hand side of an assignment is an allocatable object, it should be reallocated to the shape of the right-hand side of the assignment before the assignment occurs. This is the current Fortran Standard definition. This feature may cause extra overhead at run time. This option has the same effect as option assume realloc_lhs.

    If you specify nostandard-realloc-lhs, the compiler uses the old Fortran 2003 rules when interpreting assignment statements. The left-hand side is assumed to be allocated with the correct shape to hold the right-hand side. If it is not, incorrect behavior will occur. This option has the same effect as option assume norealloc_lhs.

• • -align array32byte
  • 
  • EXTRA_FOPTIMIZE

  • The align toggle changes how data elements are aligned. Variables and arrays are analyzed and memory layout can be altered. Specifying array32byte will look for opportunities to transform and reailgn arrays to 32byte boundaries.

627.cam4_s

• • -xCORE-AVX512
  • 
  • COPTIMIZE, FOPTIMIZE

  • Code is optimized for Intel(R) processors with support for CORE-AVX512 instructions. The resulting code may contain unconditional use of features that are not supported on other processors. This option also enables new optimizations in addition to Intel processor-specific optimizations including advanced data layout and code restructuring optimizations to improve memory accesses for Intel processors.

    Do not use this option if you are executing a program on a processor that is not an Intel processor. If you use this option on a non-compatible processor to compile the main program (in Fortran) or the function main() in C/C++, the program will display a fatal run-time error if they are executed on unsupported processors.

• • -ipo
  • 
  • COPTIMIZE, FOPTIMIZE

  • Multi-file ip optimizations that includes:
    - inline function expansion
    - interprocedural constant propogation
    - dead code elimination
    - propagation of function characteristics
    - passing arguments in registers
    - loop-invariant code motion

• • -O3
  • 
  • COPTIMIZE, FOPTIMIZE

  • Enables O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enables optimizations for maximum speed, such as:

    • Loop unrolling, including instruction scheduling
    • Code replication to eliminate branches
    • Padding the size of certain power-of-two arrays to allow more efficient cache use.
    
    

    On IA-32 and Intel EM64T processors, when O3 is used with options -ax or -x (Linux) or with options /Qax or /Qx (Windows), the compiler performs more aggressive data dependency analysis than for O2, which may result in longer compilation times. The O3 optimizations may not cause higher performance unless loop and memory access transformations take place. The optimizations may slow down code in some cases compared to O2 optimizations. The O3 option is recommended for applications that have loops that heavily use floating-point calculations and process large data sets.

  • Includes:
    • -O2
      • -O1
        • -funroll-loops
        • -fno-builtin
        • -mno-ieee-fp
        • -fomit-framepointer
        • -ffunction-sections
        • -ftz
• • -no-prec-div
  • 
  • COPTIMIZE, FOPTIMIZE
  • (disable/enable[default] -prec-div)

    -no-prec-div enables optimizations that give slightly less precise results than full IEEE division.

    When you specify -no-prec-div along with some optimizations, such as -xN and -xB (Linux) or /QxN and /QxB (Windows), the compiler may change floating-point division computations into multiplication by the reciprocal of the denominator. For example, A/B is computed as A * (1/B) to improve the speed of the computation.

    However, sometimes the value produced by this transformation is not as accurate as full IEEE division. When it is important to have fully precise IEEE division, do not use -no-prec-div. This will enable the default -prec-div and the result will be more accurate, with some loss of performance.

• • -qopt-prefetch
  • 
  • COPTIMIZE, FOPTIMIZE

  • Enable/disable(DEFAULT) the compiler to generate prefetch instructions to prefetch data.

• • -ffinite-math-only
  • 
  • COPTIMIZE, FOPTIMIZE

  • Allow optimizations for floating point arithmetic that assume arguments and results are not NaNs or Infinities

• • -qopt-mem-layout-trans=3
  • 
  • COPTIMIZE, FOPTIMIZE

  • Controls the level of memory layout transformations performed by the compiler. This option can improve cache reuse and cache locality.

    • 0: Disables memory layout transformations. This is the same as specifying -qno-opt-mem-layout-trans
    • 1: Enables basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enables more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Compiler is more aggressive in using memory layout transformations. You should only use this setting if your system has more than 4GB of physical memory per core.
• • -qopenmp
  • 
  • Yes
  • COPTIMIZE, FOPTIMIZE

• • -DSPEC_OPENMP
  • 
  • COPTIMIZE, FOPTIMIZE

  • Definition of this macro indicates that compilation for parallel operation is enabled, and that any OpenMP directives or pragmas will be visible to the compiler. The behavior of this macro is overridden if -DSPEC_SUPPRESS_OPENMP also appears in the list of compilation flags.

• • -nostandard-realloc-lhs
  • 
  • EXTRA_FOPTIMIZE

  • Option standard-realloc-lhs (the default), tells the compiler that when the left-hand side of an assignment is an allocatable object, it should be reallocated to the shape of the right-hand side of the assignment before the assignment occurs. This is the current Fortran Standard definition. This feature may cause extra overhead at run time. This option has the same effect as option assume realloc_lhs.

    If you specify nostandard-realloc-lhs, the compiler uses the old Fortran 2003 rules when interpreting assignment statements. The left-hand side is assumed to be allocated with the correct shape to hold the right-hand side. If it is not, incorrect behavior will occur. This option has the same effect as option assume norealloc_lhs.

• • -align array32byte
  • 
  • EXTRA_FOPTIMIZE

  • The align toggle changes how data elements are aligned. Variables and arrays are analyzed and memory layout can be altered. Specifying array32byte will look for opportunities to transform and reailgn arrays to 32byte boundaries.

628.pop2_s

• • Same as 621.wrf_s

Benchmarks using Fortran, C, and C++

• • -prof-gen
  • 
  • PASS1_CFLAGS, PASS1_CXXFLAGS, PASS1_FFLAGS, PASS1_LDFLAGS

  • Instrument program for profiling for the first phase of two-phase profile guided otimization. This instrumentation gathers information about a program's execution paths and data values but does not gather information from hardware performance counters. The profile instrumentation also gathers data for optimizations which are unique to profile-feedback optimization.

    -profgen:threadsafe option collects profile guided optimization data with guards for threaded applications.

• • -prof-use
  • 
  • PASS2_CFLAGS, PASS2_CXXFLAGS, PASS2_FFLAGS, PASS2_LDFLAGS

  • Instructs the compiler to produce a profile-optimized executable and merges available dynamic information (.dyn) files into a pgopti.dpi file. If you perform multiple executions of the instrumented program, -prof-use merges the dynamic information files again and overwrites the previous pgopti.dpi file.
    Without any other options, the current directory is searched for .dyn files

• • -O2
  • 
  • PASS1_COPTIMIZE, PASS1_CXXOPTIMIZE, PASS1_FOPTIMIZE

  • Enables optimizations for speed. This is the generally recommended optimization level. This option also enables:
    - Inlining of intrinsics
    - Intra-file interprocedural optimizations, which include:
    - inlining
    - constant propagation
    - forward substitution
    - routine attribute propagation
    - variable address-taken analysis
    - dead static function elimination
    - removal of unreferenced variables
    - The following capabilities for performance gain:
    - constant propagation
    - copy propagation
    - dead-code elimination
    - global register allocation
    - global instruction scheduling and control speculation
    - loop unrolling
    - optimized code selection
    - partial redundancy elimination
    - strength reduction/induction variable simplification
    - variable renaming
    - exception handling optimizations
    - tail recursions
    - peephole optimizations
    - structure assignment lowering and optimizations
    - dead store elimination
    

  • Includes:
    • -O1
      • -funroll-loops
      • -fno-builtin
      • -mno-ieee-fp
      • -fomit-framepointer
      • -ffunction-sections
      • -ftz
• • -xCORE-AVX512
  • 
  • PASS2_COPTIMIZE, PASS2_CXXOPTIMIZE, PASS2_FOPTIMIZE

  • Code is optimized for Intel(R) processors with support for CORE-AVX512 instructions. The resulting code may contain unconditional use of features that are not supported on other processors. This option also enables new optimizations in addition to Intel processor-specific optimizations including advanced data layout and code restructuring optimizations to improve memory accesses for Intel processors.

    Do not use this option if you are executing a program on a processor that is not an Intel processor. If you use this option on a non-compatible processor to compile the main program (in Fortran) or the function main() in C/C++, the program will display a fatal run-time error if they are executed on unsupported processors.

• • -qopt-prefetch
  • 
  • PASS1_COPTIMIZE, PASS1_CXXOPTIMIZE, PASS1_FOPTIMIZE, PASS2_COPTIMIZE, PASS2_CXXOPTIMIZE, PASS2_FOPTIMIZE

  • Enable/disable(DEFAULT) the compiler to generate prefetch instructions to prefetch data.

• • -ipo
  • 
  • PASS2_COPTIMIZE, PASS2_CXXOPTIMIZE, PASS2_FOPTIMIZE

  • Multi-file ip optimizations that includes:
    - inline function expansion
    - interprocedural constant propogation
    - dead code elimination
    - propagation of function characteristics
    - passing arguments in registers
    - loop-invariant code motion

• • -O3
  • 
  • PASS2_COPTIMIZE, PASS2_CXXOPTIMIZE, PASS2_FOPTIMIZE

  • Enables O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enables optimizations for maximum speed, such as:

    • Loop unrolling, including instruction scheduling
    • Code replication to eliminate branches
    • Padding the size of certain power-of-two arrays to allow more efficient cache use.
    
    

    On IA-32 and Intel EM64T processors, when O3 is used with options -ax or -x (Linux) or with options /Qax or /Qx (Windows), the compiler performs more aggressive data dependency analysis than for O2, which may result in longer compilation times. The O3 optimizations may not cause higher performance unless loop and memory access transformations take place. The optimizations may slow down code in some cases compared to O2 optimizations. The O3 option is recommended for applications that have loops that heavily use floating-point calculations and process large data sets.

  • Includes:
    • -O2
      • -O1
        • -funroll-loops
        • -fno-builtin
        • -mno-ieee-fp
        • -fomit-framepointer
        • -ffunction-sections
        • -ftz
• • -ffinite-math-only
  • 
  • PASS1_COPTIMIZE, PASS1_CXXOPTIMIZE, PASS1_FOPTIMIZE, PASS2_COPTIMIZE, PASS2_CXXOPTIMIZE, PASS2_FOPTIMIZE

  • Allow optimizations for floating point arithmetic that assume arguments and results are not NaNs or Infinities

• • -no-prec-div
  • 
  • PASS2_COPTIMIZE, PASS2_CXXOPTIMIZE, PASS2_FOPTIMIZE
  • (disable/enable[default] -prec-div)

    -no-prec-div enables optimizations that give slightly less precise results than full IEEE division.

    When you specify -no-prec-div along with some optimizations, such as -xN and -xB (Linux) or /QxN and /QxB (Windows), the compiler may change floating-point division computations into multiplication by the reciprocal of the denominator. For example, A/B is computed as A * (1/B) to improve the speed of the computation.

    However, sometimes the value produced by this transformation is not as accurate as full IEEE division. When it is important to have fully precise IEEE division, do not use -no-prec-div. This will enable the default -prec-div and the result will be more accurate, with some loss of performance.

• • -qopt-mem-layout-trans=3
  • 
  • PASS1_COPTIMIZE, PASS1_CXXOPTIMIZE, PASS1_FOPTIMIZE, PASS2_COPTIMIZE, PASS2_CXXOPTIMIZE, PASS2_FOPTIMIZE

  • Controls the level of memory layout transformations performed by the compiler. This option can improve cache reuse and cache locality.

    • 0: Disables memory layout transformations. This is the same as specifying -qno-opt-mem-layout-trans
    • 1: Enables basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enables more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Compiler is more aggressive in using memory layout transformations. You should only use this setting if your system has more than 4GB of physical memory per core.
• • -DSPEC_SUPPRESS_OPENMP
  • 
  • PASS1_COPTIMIZE, PASS1_CXXOPTIMIZE, PASS1_FOPTIMIZE

  • Definition of this macro indicates that compilation for parallel operation is disabled, and that any OpenMP directives or pragmas will not be visible to the compiler. This macro overrides the behavior of -DSPEC_OPENMP when they both appear in the list of compilation flags.

• • -qopenmp
  • 
  • Yes
  • PASS2_COPTIMIZE, PASS2_CXXOPTIMIZE, PASS2_FOPTIMIZE

• • -DSPEC_OPENMP
  • 
  • PASS2_COPTIMIZE, PASS2_CXXOPTIMIZE, PASS2_FOPTIMIZE

  • Definition of this macro indicates that compilation for parallel operation is enabled, and that any OpenMP directives or pragmas will be visible to the compiler. The behavior of this macro is overridden if -DSPEC_SUPPRESS_OPENMP also appears in the list of compilation flags.

• • -nostandard-realloc-lhs
  • 
  • EXTRA_FOPTIMIZE

  • Option standard-realloc-lhs (the default), tells the compiler that when the left-hand side of an assignment is an allocatable object, it should be reallocated to the shape of the right-hand side of the assignment before the assignment occurs. This is the current Fortran Standard definition. This feature may cause extra overhead at run time. This option has the same effect as option assume realloc_lhs.

    If you specify nostandard-realloc-lhs, the compiler uses the old Fortran 2003 rules when interpreting assignment statements. The left-hand side is assumed to be allocated with the correct shape to hold the right-hand side. If it is not, incorrect behavior will occur. This option has the same effect as option assume norealloc_lhs.

• • -align array32byte
  • 
  • EXTRA_FOPTIMIZE

  • The align toggle changes how data elements are aligned. Variables and arrays are analyzed and memory layout can be altered. Specifying array32byte will look for opportunities to transform and reailgn arrays to 32byte boundaries.