SPEC CPU®2026 Flag Description
Dell Inc. PowerEdge R770 (Intel Xeon 6787P)

Base Optimization Flags

C benchmarks

• • -m64
  • 
  • intel_icc,intel_icpc,intel_ifort,intel_icx,intel_icpx,intel_ifx
  • CC, LD

  • Compiles for a 64-bit (LP64) data model.

• • -std=c18
  • 
  • intel_icx,intel_icpx
  • CC, LD

  • Sets the language standard to the specified version, for example c18, c++17, f2008.

• • -Wl,-z,muldefs
  • 
  • EXTRA_LDFLAGS

  • Enable SmartHeap and/or other library usage by forcing the linker to ignore multiple definitions if present

• • -xgraniterapids
  • 
  • COPTIMIZE

  • May generate instructions for processors that support the specified Intel processor or microarchitecture code name. Optimizes for the specified Intel processor or microarchitecture code name.

• • -mprefer-vector-width=512
  • 
  • COPTIMIZE

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

• • -O3
  • 
  • COPTIMIZE

  • Enable O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enable 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
• • -ffp-model=fast
  • 
  • COPTIMIZE

  • Tells the compiler to use more aggressive optimizations when implementing floating-point calculations. These optimizations increase speed, but they may affect the accuracy or reproducibility of floating-point computations.

    Specifying fast is the same as specifying fast=1. Setting fast=1 (the default) recognizes and supports NaN and infinite values.

    For Intel compiler options ffp-model=fast and fp-model=fast are interchangeable.

• • -flto
  • 
  • COPTIMIZE

  • Performs link time optimizations, which is also known as Interprocedural Optimizations.

• • -mfpmath=sse
  • 
  • COPTIMIZE

  • Generate floating-point arithmetic for selected unit unit. Here use scalar floating-point instructions present in the SSE instruction set

• • -funroll-loops
  • 
  • COPTIMIZE

  • Tells the compiler the maximum number of times to unroll loops. For example -funroll-loops0 would disable unrolling of loops.

• • -qopt-mem-layout-trans=4
  • 
  • 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: Enable basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enable more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Enable more memory layout transformations like copy-in/copy-out of structures for a region of code. You should only use this setting if your system has more than 4GB of physical memory per core.
    • 4: 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.
• • -L/usr/local/jemalloc-5.3.0/lib
  • 
  • EXTRA_LIBS

  • Build time link path for libraries supplied with the compiler (for example, the qkmalloc library).

• • -ljemalloc
  • 
  • EXTRA_LIBS

  • Linker toggle to specify jemalloc linker library. See jemalloc.net for more information.

C++ benchmarks

• • -m64
  • 
  • intel_icc,intel_icpc,intel_ifort,intel_icx,intel_icpx,intel_ifx
  • CXX, LD

  • Compiles for a 64-bit (LP64) data model.

• • -std=c++17
  • 
  • intel_icx,intel_icpx
  • CXX, LD

  • Sets the language standard to the specified version, for example c18, c++17, f2008.

• • -Wl,-z,muldefs
  • 
  • EXTRA_LDFLAGS

  • Enable SmartHeap and/or other library usage by forcing the linker to ignore multiple definitions if present

• • -xgraniterapids
  • 
  • CXXOPTIMIZE

  • May generate instructions for processors that support the specified Intel processor or microarchitecture code name. Optimizes for the specified Intel processor or microarchitecture code name.

• • -mprefer-vector-width=512
  • 
  • CXXOPTIMIZE

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

• • -O3
  • 
  • CXXOPTIMIZE

  • Enable O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enable 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
• • -ffp-model=fast
  • 
  • CXXOPTIMIZE

  • Tells the compiler to use more aggressive optimizations when implementing floating-point calculations. These optimizations increase speed, but they may affect the accuracy or reproducibility of floating-point computations.

    Specifying fast is the same as specifying fast=1. Setting fast=1 (the default) recognizes and supports NaN and infinite values.

    For Intel compiler options ffp-model=fast and fp-model=fast are interchangeable.

• • -flto
  • 
  • CXXOPTIMIZE

  • Performs link time optimizations, which is also known as Interprocedural Optimizations.

• • -mfpmath=sse
  • 
  • CXXOPTIMIZE

  • Generate floating-point arithmetic for selected unit unit. Here use scalar floating-point instructions present in the SSE instruction set

• • -funroll-loops
  • 
  • CXXOPTIMIZE

  • Tells the compiler the maximum number of times to unroll loops. For example -funroll-loops0 would disable unrolling of loops.

• • -qopt-mem-layout-trans=4
  • 
  • CXXOPTIMIZE

  • 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: Enable basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enable more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Enable more memory layout transformations like copy-in/copy-out of structures for a region of code. You should only use this setting if your system has more than 4GB of physical memory per core.
    • 4: 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.
• • -L/usr/local/jemalloc-5.3.0/lib
  • 
  • EXTRA_LIBS

  • Build time link path for libraries supplied with the compiler (for example, the qkmalloc library).

• • -ljemalloc
  • 
  • EXTRA_LIBS

  • Linker toggle to specify jemalloc linker library. See jemalloc.net for more information.

Fortran benchmarks

• • -m64
  • 
  • intel_icc,intel_icpc,intel_ifort,intel_icx,intel_icpx,intel_ifx
  • FC, LD

  • Compiles for a 64-bit (LP64) data model.

• • -stand f18
  • 
  • intel_ifx
  • FC, LD

  • Sets the language dialect to conform to the indicated Fortran standard.

• • -Wl,-z,muldefs
  • 
  • EXTRA_LDFLAGS

  • Enable SmartHeap and/or other library usage by forcing the linker to ignore multiple definitions if present

• • -xgraniterapids
  • 
  • FOPTIMIZE

  • May generate instructions for processors that support the specified Intel processor or microarchitecture code name. Optimizes for the specified Intel processor or microarchitecture code name.

• • -mprefer-vector-width=512
  • 
  • FOPTIMIZE

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

• • -O3
  • 
  • FOPTIMIZE

  • Enable O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enable 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
• • -ffp-model=fast
  • 
  • FOPTIMIZE

  • Tells the compiler to use more aggressive optimizations when implementing floating-point calculations. These optimizations increase speed, but they may affect the accuracy or reproducibility of floating-point computations.

    Specifying fast is the same as specifying fast=1. Setting fast=1 (the default) recognizes and supports NaN and infinite values.

    For Intel compiler options ffp-model=fast and fp-model=fast are interchangeable.

• • -flto
  • 
  • FOPTIMIZE

  • Performs link time optimizations, which is also known as Interprocedural Optimizations.

• • -mfpmath=sse
  • 
  • FOPTIMIZE

  • Generate floating-point arithmetic for selected unit unit. Here use scalar floating-point instructions present in the SSE instruction set

• • -funroll-loops
  • 
  • FOPTIMIZE

  • Tells the compiler the maximum number of times to unroll loops. For example -funroll-loops0 would disable unrolling of loops.

• • -qopt-mem-layout-trans=4
  • 
  • 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: Enable basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enable more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Enable more memory layout transformations like copy-in/copy-out of structures for a region of code. You should only use this setting if your system has more than 4GB of physical memory per core.
    • 4: 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.
• • -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.

• • -auto
  • 
  • EXTRA_FOPTIMIZE

  • Make all local variables AUTOMATIC. Same as -automatic

• • -L/usr/local/jemalloc-5.3.0/lib
  • 
  • EXTRA_LIBS

  • Build time link path for libraries supplied with the compiler (for example, the qkmalloc library).

• • -ljemalloc
  • 
  • EXTRA_LIBS

  • Linker toggle to specify jemalloc linker library. See jemalloc.net for more information.

Benchmarks using both C and C++

• • -m64
  • 
  • intel_icc,intel_icpc,intel_ifort,intel_icx,intel_icpx,intel_ifx
  • CC, CXX, LD

  • Compiles for a 64-bit (LP64) data model.

• • -std=c++17
  • 
  • intel_icx,intel_icpx
  • CXX, LD

  • Sets the language standard to the specified version, for example c18, c++17, f2008.

• • -std=c18
  • 
  • intel_icx,intel_icpx
  • CC

  • Sets the language standard to the specified version, for example c18, c++17, f2008.

• • -Wl,-z,muldefs
  • 
  • EXTRA_LDFLAGS

  • Enable SmartHeap and/or other library usage by forcing the linker to ignore multiple definitions if present

• • -xgraniterapids
  • 
  • COPTIMIZE, CXXOPTIMIZE

  • May generate instructions for processors that support the specified Intel processor or microarchitecture code name. Optimizes for the specified Intel processor or microarchitecture code name.

• • -mprefer-vector-width=512
  • 
  • COPTIMIZE, CXXOPTIMIZE

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

• • -O3
  • 
  • COPTIMIZE, CXXOPTIMIZE

  • Enable O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enable 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
• • -ffp-model=fast
  • 
  • COPTIMIZE, CXXOPTIMIZE

  • Tells the compiler to use more aggressive optimizations when implementing floating-point calculations. These optimizations increase speed, but they may affect the accuracy or reproducibility of floating-point computations.

    Specifying fast is the same as specifying fast=1. Setting fast=1 (the default) recognizes and supports NaN and infinite values.

    For Intel compiler options ffp-model=fast and fp-model=fast are interchangeable.

• • -flto
  • 
  • COPTIMIZE, CXXOPTIMIZE

  • Performs link time optimizations, which is also known as Interprocedural Optimizations.

• • -mfpmath=sse
  • 
  • COPTIMIZE, CXXOPTIMIZE

  • Generate floating-point arithmetic for selected unit unit. Here use scalar floating-point instructions present in the SSE instruction set

• • -funroll-loops
  • 
  • COPTIMIZE, CXXOPTIMIZE

  • Tells the compiler the maximum number of times to unroll loops. For example -funroll-loops0 would disable unrolling of loops.

• • -qopt-mem-layout-trans=4
  • 
  • COPTIMIZE, CXXOPTIMIZE

  • 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: Enable basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enable more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Enable more memory layout transformations like copy-in/copy-out of structures for a region of code. You should only use this setting if your system has more than 4GB of physical memory per core.
    • 4: 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.
• • -L/usr/local/jemalloc-5.3.0/lib
  • 
  • EXTRA_LIBS

  • Build time link path for libraries supplied with the compiler (for example, the qkmalloc library).

• • -ljemalloc
  • 
  • EXTRA_LIBS

  • Linker toggle to specify jemalloc linker library. See jemalloc.net for more information.

Peak Optimization Flags

C benchmarks

782.lbm_r

• • basepeak = yes

C++ benchmarks

731.astcenc_r

• • -m64
  • 
  • intel_icc,intel_icpc,intel_ifort,intel_icx,intel_icpx,intel_ifx
  • CXX, LD

  • Compiles for a 64-bit (LP64) data model.

• • -std=c++17
  • 
  • intel_icx,intel_icpx
  • CXX, LD

  • Sets the language standard to the specified version, for example c18, c++17, f2008.

• • -Wl,-z,muldefs
  • 
  • EXTRA_LDFLAGS

  • Enable SmartHeap and/or other library usage by forcing the linker to ignore multiple definitions if present

• • -fprofile-generate
  • 
  • PASS1_CXXFLAGS, 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.

• • -fprofile-use=default.profdata
  • 
  • PASS2_CXXFLAGS, PASS2_LDFLAGS

  • Instructs the compiler to produce a profile-optimized executable and merges available dynamic information.

• • -xHost
  • 
  • PASS1_CXXFLAGS, PASS1_LDFLAGS

  • This option tells the compiler to generate instructions for the highest instruction set available on the compilation host processor. The instructions generated by Host differ depending on the compilation host processor.

• • -ffp-model=fast
  • 
  • PASS1_CXXFLAGS, PASS1_LDFLAGS, PASS2_CXXFLAGS, PASS2_LDFLAGS

  • Tells the compiler to use more aggressive optimizations when implementing floating-point calculations. These optimizations increase speed, but they may affect the accuracy or reproducibility of floating-point computations.

    Specifying fast is the same as specifying fast=1. Setting fast=1 (the default) recognizes and supports NaN and infinite values.

    For Intel compiler options ffp-model=fast and fp-model=fast are interchangeable.

• • -xgraniterapids
  • 
  • PASS2_CXXFLAGS, PASS2_LDFLAGS

  • May generate instructions for processors that support the specified Intel processor or microarchitecture code name. Optimizes for the specified Intel processor or microarchitecture code name.

• • -flto
  • 
  • PASS1_CXXFLAGS, PASS1_LDFLAGS, PASS2_CXXFLAGS, PASS2_LDFLAGS

  • Performs link time optimizations, which is also known as Interprocedural Optimizations.

• • -mprefer-vector-width=512
  • 
  • PASS1_CXXFLAGS, PASS1_LDFLAGS, PASS2_CXXFLAGS, PASS2_LDFLAGS

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

• • -qopt-mem-layout-trans=4
  • 
  • PASS1_CXXFLAGS, PASS1_LDFLAGS, PASS2_CXXFLAGS, PASS2_LDFLAGS

  • 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: Enable basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enable more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Enable more memory layout transformations like copy-in/copy-out of structures for a region of code. You should only use this setting if your system has more than 4GB of physical memory per core.
    • 4: 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.
• • -O3
  • 
  • PASS1_CXXFLAGS, PASS2_CXXFLAGS

  • Enable O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enable 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
• • -mfpmath=sse
  • 
  • PASS1_CXXFLAGS, PASS2_CXXFLAGS

  • Generate floating-point arithmetic for selected unit unit. Here use scalar floating-point instructions present in the SSE instruction set

• • -funroll-loops
  • 
  • PASS1_CXXFLAGS, PASS2_CXXFLAGS

  • Tells the compiler the maximum number of times to unroll loops. For example -funroll-loops0 would disable unrolling of loops.

• • -L/usr/local/jemalloc-5.3.0/lib
  • 
  • EXTRA_LIBS

  • Build time link path for libraries supplied with the compiler (for example, the qkmalloc library).

• • -ljemalloc
  • 
  • EXTRA_LIBS

  • Linker toggle to specify jemalloc linker library. See jemalloc.net for more information.

736.ocio_r

• • basepeak = yes

748.flightdm_r

• • Same as 731.astcenc_r

766.femflow_r

• • basepeak = yes

767.nest_r

• • Same as 731.astcenc_r

772.marian_r

• • basepeak = yes

Fortran benchmarks

722.palm_r

• • basepeak = yes

749.fotonik3d_r

• • basepeak = yes

765.roms_r

• • basepeak = yes

Benchmarks using both C and C++

709.cactus_r

• • basepeak = yes

737.gmsh_r

• • -m64
  • 
  • intel_icc,intel_icpc,intel_ifort,intel_icx,intel_icpx,intel_ifx
  • CC, CXX, LD

  • Compiles for a 64-bit (LP64) data model.

• • -std=c++17
  • 
  • intel_icx,intel_icpx
  • CXX, LD

  • Sets the language standard to the specified version, for example c18, c++17, f2008.

• • -std=c18
  • 
  • intel_icx,intel_icpx
  • CC

  • Sets the language standard to the specified version, for example c18, c++17, f2008.

• • -Wl,-z,muldefs
  • 
  • EXTRA_LDFLAGS

  • Enable SmartHeap and/or other library usage by forcing the linker to ignore multiple definitions if present

• • -fprofile-generate
  • 
  • PASS1_CFLAGS, PASS1_CXXFLAGS, 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.

• • -fprofile-use=default.profdata
  • 
  • PASS2_CFLAGS, PASS2_CXXFLAGS, PASS2_LDFLAGS

  • Instructs the compiler to produce a profile-optimized executable and merges available dynamic information.

• • -xHost
  • 
  • PASS1_CFLAGS, PASS1_CXXFLAGS, PASS1_LDFLAGS

  • This option tells the compiler to generate instructions for the highest instruction set available on the compilation host processor. The instructions generated by Host differ depending on the compilation host processor.

• • -ffp-model=fast
  • 
  • PASS1_CFLAGS, PASS1_CXXFLAGS, PASS1_LDFLAGS, PASS2_CFLAGS, PASS2_CXXFLAGS, PASS2_LDFLAGS

  • Tells the compiler to use more aggressive optimizations when implementing floating-point calculations. These optimizations increase speed, but they may affect the accuracy or reproducibility of floating-point computations.

    Specifying fast is the same as specifying fast=1. Setting fast=1 (the default) recognizes and supports NaN and infinite values.

    For Intel compiler options ffp-model=fast and fp-model=fast are interchangeable.

• • -xgraniterapids
  • 
  • PASS2_CFLAGS, PASS2_CXXFLAGS, PASS2_LDFLAGS

  • May generate instructions for processors that support the specified Intel processor or microarchitecture code name. Optimizes for the specified Intel processor or microarchitecture code name.

• • -flto
  • 
  • PASS1_CFLAGS, PASS1_CXXFLAGS, PASS1_LDFLAGS, PASS2_CFLAGS, PASS2_CXXFLAGS, PASS2_LDFLAGS

  • Performs link time optimizations, which is also known as Interprocedural Optimizations.

• • -mprefer-vector-width=512
  • 
  • PASS1_CFLAGS, PASS1_CXXFLAGS, PASS1_LDFLAGS, PASS2_CFLAGS, PASS2_CXXFLAGS, PASS2_LDFLAGS

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

• • -qopt-mem-layout-trans=4
  • 
  • PASS1_CFLAGS, PASS1_CXXFLAGS, PASS1_LDFLAGS, PASS2_CFLAGS, PASS2_CXXFLAGS, PASS2_LDFLAGS

  • 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: Enable basic memory layout transformations like structure splitting, structure peeling, field inlining, field reordering, array field transpose, increase field alignment etc.
    • 2: Enable more memory layout transformations like advanced structure splitting. This is the same as specifying -qopt-mem-layout-trans
    • 3: Enable more memory layout transformations like copy-in/copy-out of structures for a region of code. You should only use this setting if your system has more than 4GB of physical memory per core.
    • 4: 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.
• • -O3
  • 
  • PASS1_CFLAGS, PASS1_CXXFLAGS, PASS2_CFLAGS, PASS2_CXXFLAGS

  • Enable O2 optimizations plus more aggressive optimizations, such as prefetching, scalar replacement, and loop and memory access transformations. Enable 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
• • -mfpmath=sse
  • 
  • PASS1_CFLAGS, PASS1_CXXFLAGS, PASS2_CFLAGS, PASS2_CXXFLAGS

  • Generate floating-point arithmetic for selected unit unit. Here use scalar floating-point instructions present in the SSE instruction set

• • -funroll-loops
  • 
  • PASS1_CFLAGS, PASS1_CXXFLAGS, PASS2_CFLAGS, PASS2_CXXFLAGS

  • Tells the compiler the maximum number of times to unroll loops. For example -funroll-loops0 would disable unrolling of loops.

• • -L/usr/local/jemalloc-5.3.0/lib
  • 
  • EXTRA_LIBS

  • Build time link path for libraries supplied with the compiler (for example, the qkmalloc library).

• • -ljemalloc
  • 
  • EXTRA_LIBS

  • Linker toggle to specify jemalloc linker library. See jemalloc.net for more information.