Skip to main content

SPEC CPU Profiles

The following profiles run customer-representative or benchmarking scenarios using the SPEC CPU workload. Ensure that you have read the information on getting and integrating SPEC CPU into Virtual Client in the workload details documentation linked below.

info

SPEC CPU workloads are long-running workloads. When running in Peak performance scenarios, some take more than 2 days to complete. Check the 'Profile Runtimes' section below for more details on what to expect.

PERF-SPECCPU-FPRATE.json​

Runs the SPEC CPU 2017 Floating Point Rate (fprate) benchmark workload on the system focusing on baseline + peak measurements. This workload is an industry standard for evaluating the performance of the CPU for processing calculations.

  • Workload Profile

  • Supported Platform-Architectures

    • linux-x64
    • linux-arm64
    • win-x64
    • win-arm64

  • Dependencies
    The dependencies defined in the 'Dependencies' section of the profile itself are required in order to run the workload operations effectively.

    • Internet connection.

    Additional information on components that exist within the 'Dependencies' section of the profile can be found in the following locations:

  • Profile Parameters
    The following parameters can be optionally supplied on the command line to modify the behaviors of the workload.

    ParameterPurposeDefault value
    CompilerVersionOptional. The version of the compiler to use.13
    IterationCountOptional. The number of iterations for speccpu to run the benchmark/suite.2
    BenchmarksOptional. The benchmark suite or list of benchmarks to run.fprate
    RunPeakOptional. True to run the workload 'Peak' scenario, False to run the workload 'Base' scenario.false (Base)
    ThreadCountOptional. Determines the number of threads to use for running the benchmark.# logical processors
    CopyCountOptional. Determines the number of copies of the benchmark to run concurrently.# logical processors
    BaseOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Base' scenario
    • X64 systems: -g -O2 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O2 -march=armv8.2-a -flto -frecord-gcc-switches
    PeakOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Peak' scenario
    • X64 systems: -g -O3 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O3 -march=armv8.2-a -flto -frecord-gcc-switches

  • Profile Runtimes
    The following timings represent the length of time required to run a single round of profile actions. These timings can be used to determine minimum required runtimes for the Virtual Client in order to get results. These are estimates based on the number of system cores. When running the SPEC CPU workload, the timing can be difficult to predict. It is recommended that you avoid using timeouts on the command line and use iterations instead (e.g. --timeout=1440 -> --iterations=1)

    • Base Scenario (8-cores/vCPUs) =~ 10 hours
    • Base scenario (48-cores/vCPUs) =~ 14 hours
    • Peak scenario (8-cores/vCPUs) =~ 18 hours
    • Peak scenario (48-cores/vCPUs) =~ 28 hours

  • Usage Examples
    The following section provides a few basic examples of how to use the workload profile. Additional usage examples can be found in the 'Usage Scenarios/Examples' link at the top.

  • Notes
    On Windows, only the following benchmarks are supported: 503.bwaves_r, 508.namd_r, 510.parest_r, 519.lbm_r, 544.nab_r, 549.fotonik3d_r, 554.roms_r.

    # Execute the workload profile
    ./VirtualClient --profile=PERF-SPECCPU-FPRATE.json --iterations=1 --logger=csv

PERF-SPECCPU-FPSPEED.json​

Runs the SPEC CPU 2017 Floating Point Speed (fpspeed) benchmark workload on the system focusing on baseline + peak measurements. This workload is an industry standard for evaluating the performance of the CPU for processing calculations.

  • Workload Profile

  • OS/Architecture Platforms

    • linux-x64
    • linux-arm64
    • win-x64
    • win-arm64

  • Dependencies
    The dependencies defined in the 'Dependencies' section of the profile itself are required in order to run the workload operations effectively.

    • Internet connection.

    Additional information on components that exist within the 'Dependencies' section of the profile can be found in the following locations:

  • Profile Parameters
    The following parameters can be optionally supplied on the command line to modify the behaviors of the workload.

    ParameterPurposeDefault value
    CompilerVersionOptional. The version of the compiler to use.13
    IterationCountOptional. The number of iterations for speccpu to run the benchmark/suite.2
    BenchmarksOptional. The benchmark suite or list of benchmarks to run.fpspeed
    RunPeakOptional. True to run the workload 'Peak' scenario, False to run the workload 'Base' scenario.false (Base)
    ThreadCountOptional. Determines the number of threads to use for running the benchmark.# logical processors
    CopyCountOptional. Determines the number of copies of the benchmark to run concurrently.# logical processors
    BaseOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Base' scenario
    • X64 systems: -g -O2 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O2 -march=armv8.2-a -flto -frecord-gcc-switches
    PeakOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Peak' scenario
    • X64 systems: -g -O3 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O3 -march=armv8.2-a -flto -frecord-gcc-switches

  • Profile Runtimes
    See the 'Metadata' section of the profile for estimated runtimes. These timings represent the length of time required to run a single round of profile actions. These timings can be used to determine minimum required runtimes for the Virtual Client in order to get results. These are often estimates based on the number of system cores. When running the SPEC CPU workload, the timing can be difficult to predict. It is recommended that you avoid using timeouts on the command line and use iterations instead (e.g. --timeout=1440 -> --iterations=1)

    • Base Scenario (8-cores/vCPUs) =~ 4 hours
    • Base scenario (48-cores/vCPUs) =~ 6 hours
    • Peak scenario (8-cores/vCPUs) =~ 5 hours
    • Peak scenario (48-cores/vCPUs) =~ 8 hours

  • Usage Examples
    The following section provides a few basic examples of how to use the workload profile.

  • Notes
    On Windows, only the following benchmarks are supported: 603.bwaves_s, 619.lbm_s, 644.nab_s, 654.roms_s.

    # Execute the workload profile
    ./VirtualClient --profile=PERF-SPECCPU-FPSPEED.json --iterations=1 --logger=csv

PERF-SPECCPU-INTRATE.json​

Runs the SPEC CPU 2017 Integer Rate (intrate) benchmark workload on the system focusing on baseline + peak measurements. This workload is an industry standard for evaluating the performance of the CPU for processing calculations.

  • Workload Profile

  • OS/Architecture Platforms

    • linux-x64
    • linux-arm64
    • win-x64
    • win-arm64

  • Dependencies
    The dependencies defined in the 'Dependencies' section of the profile itself are required in order to run the workload operations effectively.

    • Internet connection.

    Additional information on components that exist within the 'Dependencies' section of the profile can be found in the following locations:

  • Profile Parameters
    The following parameters can be optionally supplied on the command line to modify the behaviors of the workload.

    ParameterPurposeDefault value
    CompilerVersionOptional. The version of the compiler to use.13
    IterationCountOptional. The number of iterations for speccpu to run the benchmark/suite.2
    BenchmarksOptional. The benchmark suite or list of benchmarks to run.intrate
    RunPeakOptional. True to run the workload 'Peak' scenario, False to run the workload 'Base' scenario.false (Base)
    ThreadCountOptional. Determines the number of threads to use for running the benchmark.# logical processors
    CopyCountOptional. Determines the number of copies of the benchmark to run concurrently.# logical processors
    BaseOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Base' scenario
    • X64 systems: -g -O2 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O2 -march=armv8.2-a -flto -frecord-gcc-switches
    PeakOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Peak' scenario
    • X64 systems: -g -O3 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O3 -march=armv8.2-a -flto -frecord-gcc-switches

  • Profile Runtimes
    See the 'Metadata' section of the profile for estimated runtimes. These timings represent the length of time required to run a single round of profile actions. These timings can be used to determine minimum required runtimes for the Virtual Client in order to get results. These are often estimates based on the number of system cores. When running the SPEC CPU workload, the timing can be difficult to predict. It is recommended that you avoid using timeouts on the command line and use iterations instead (e.g. --timeout=1440 -> --iterations=1)

    • Base Scenario (8-cores/vCPUs) =~ 10 hours
    • Base scenario (48-cores/vCPUs) =~ 14 hours
    • Peak scenario (8-cores/vCPUs) =~ 18 hours
    • Peak scenario (48-cores/vCPUs) =~ 28 hours

  • Usage Examples
    The following section provides a few basic examples of how to use the workload profile.

  • Notes
    On Windows, only the following benchmarks are supported: 505.mcf_r, 525.x264_r, 541.leela_r, 548.exchange2_r, 557.xz_r.

    # Execute the workload profile
    ./VirtualClient --profile=PERF-SPECCPU-INTRATE.json --iterations=1 --logger=csv

PERF-SPECCPU-INTSPEED.json​

Runs the SPEC CPU 2017 Integer Speed (intspeed) benchmark workload on the system focusing on baseline + peak measurements. This workload is an industry standard for evaluating the performance of the CPU for processing calculations.

  • Workload Profile

  • OS/Architecture Platforms

    • linux-x64
    • linux-arm64
    • win-x64
    • win-arm64

  • Dependencies
    The dependencies defined in the 'Dependencies' section of the profile itself are required in order to run the workload operations effectively.

    • Internet connection.

    Additional information on components that exist within the 'Dependencies' section of the profile can be found in the following locations:

  • Profile Parameters
    The following parameters can be optionally supplied on the command line to modify the behaviors of the workload.

    ParameterPurposeDefault value
    CompilerVersionOptional. The version of the compiler to use.13
    IterationCountOptional. The number of iterations for speccpu to run the benchmark/suite.2
    BenchmarksOptional. The benchmark suite or list of benchmarks to run.intspeed
    RunPeakOptional. True to run the workload 'Peak' scenario, False to run the workload 'Base' scenario.false (Base)
    ThreadCountOptional. Determines the number of threads to use for running the benchmark.# logical processors
    CopyCountOptional. Determines the number of copies of the benchmark to run concurrently.# logical processors
    BaseOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Base' scenario
    • X64 systems: -g -O2 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O2 -march=armv8.2-a -flto -frecord-gcc-switches
    PeakOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Peak' scenario
    • X64 systems: -g -O3 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O3 -march=armv8.2-a -flto -frecord-gcc-switches

  • Profile Runtimes
    See the 'Metadata' section of the profile for estimated runtimes. These timings represent the length of time required to run a single round of profile actions. These timings can be used to determine minimum required runtimes for the Virtual Client in order to get results. These are often estimates based on the number of system cores. When running the SPEC CPU workload, the timing can be difficult to predict. It is recommended that you avoid using timeouts on the command line and use iterations instead (e.g. --timeout=1440 -> --iterations=1)

    • Base Scenario (8-cores/vCPUs) =~ 4 hours
    • Base scenario (48-cores/vCPUs) =~ 6 hours
    • Peak scenario (8-cores/vCPUs) =~ 5 hours
    • Peak scenario (48-cores/vCPUs) =~ 8 hours

  • Usage Examples
    The following section provides a few basic examples of how to use the workload profile.

  • Notes
    On Windows, only the following benchmarks are supported: 641.leela_s.

    # Execute the workload profile
    ./VirtualClient --profile=PERF-SPECCPU-INTSPEED.json --iterations=1 --logger=csv

PERF-SPECCPU2026-FPRATE.json​

Runs the SPEC CPU 2026 Floating Point Rate (fprate) benchmark workload on the system focusing on baseline + peak measurements. This workload is an industry standard for evaluating the performance of the CPU for processing calculations.

  • Workload Profile

  • Supported Platform-Architectures

    • linux-x64
    • linux-arm64

    Note: On Windows platform, the gcc compiler version is required to be 10 or higher.

  • Dependencies
    The dependencies defined in the 'Dependencies' section of the profile itself are required in order to run the workload operations effectively.

    • Internet connection.

    Additional information on components that exist within the 'Dependencies' section of the profile can be found in the following locations:

  • Profile Parameters
    The following parameters can be optionally supplied on the command line to modify the behaviors of the workload.

    ParameterPurposeDefault value
    CompilerVersionOptional. The version of the compiler to use.13
    IterationCountOptional. The number of iterations for speccpu to run the benchmark/suite.2
    BenchmarksOptional. The benchmark suite or list of benchmarks to run.fprate
    RunPeakOptional. True to run the workload 'Peak' scenario, False to run the workload 'Base' scenario.false (Base)
    ThreadCountOptional. Determines the number of threads to use for running the benchmark.# logical processors
    CopyCountOptional. Determines the number of copies of the benchmark to run concurrently.# logical processors
    BaseOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Base' scenario
    • X64 systems: -g -O2 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O2 -march=armv8.2-a -flto -frecord-gcc-switches
    PeakOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Peak' scenario
    • X64 systems: -g -O3 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O3 -march=armv8.2-a -flto -frecord-gcc-switches

  • Profile Runtimes
    The following timings represent the length of time required to run a single round of profile actions. These timings can be used to determine minimum required runtimes for the Virtual Client in order to get results. These are estimates based on the number of system cores. When running the SPEC CPU workload, the timing can be difficult to predict. It is recommended that you avoid using timeouts on the command line and use iterations instead (e.g. --timeout=1440 -> --iterations=1)

    • Base Scenario (8-cores/vCPUs) =~ 10 hours
    • Base scenario (48-cores/vCPUs) =~ 14 hours
    • Peak scenario (8-cores/vCPUs) =~ 18 hours
    • Peak scenario (48-cores/vCPUs) =~ 28 hours

  • Usage Examples
    The following section provides a few basic examples of how to use the workload profile. Additional usage examples can be found in the 'Usage Scenarios/Examples' link at the top.

    # Execute the workload profile
    ./VirtualClient --profile=PERF-SPECCPU2026-FPRATE.json --iterations=1 --logger=csv

PERF-SPECCPU2026-FPSPEED.json​

Runs the SPEC CPU 2026 Floating Point Speed (fpspeed) benchmark workload on the system focusing on baseline + peak measurements. This workload is an industry standard for evaluating the performance of the CPU for processing calculations.

  • Workload Profile

  • OS/Architecture Platforms

    • linux-x64
    • linux-arm64

  • Dependencies
    The dependencies defined in the 'Dependencies' section of the profile itself are required in order to run the workload operations effectively.

    • Internet connection.

    Additional information on components that exist within the 'Dependencies' section of the profile can be found in the following locations:

  • Profile Parameters
    The following parameters can be optionally supplied on the command line to modify the behaviors of the workload.

    ParameterPurposeDefault value
    CompilerVersionOptional. The version of the compiler to use.13
    IterationCountOptional. The number of iterations for speccpu to run the benchmark/suite.2
    BenchmarksOptional. The benchmark suite or list of benchmarks to run.fpspeed
    RunPeakOptional. True to run the workload 'Peak' scenario, False to run the workload 'Base' scenario.false (Base)
    ThreadCountOptional. Determines the number of threads to use for running the benchmark.# logical processors
    CopyCountOptional. Determines the number of copies of the benchmark to run concurrently.# logical processors
    BaseOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Base' scenario
    • X64 systems: -g -O2 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O2 -march=armv8.2-a -flto -frecord-gcc-switches
    PeakOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Peak' scenario
    • X64 systems: -g -O3 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O3 -march=armv8.2-a -flto -frecord-gcc-switches

  • Profile Runtimes
    See the 'Metadata' section of the profile for estimated runtimes. These timings represent the length of time required to run a single round of profile actions. These timings can be used to determine minimum required runtimes for the Virtual Client in order to get results. These are often estimates based on the number of system cores. When running the SPEC CPU workload, the timing can be difficult to predict. It is recommended that you avoid using timeouts on the command line and use iterations instead (e.g. --timeout=1440 -> --iterations=1)

    • Base Scenario (8-cores/vCPUs) =~ 4 hours
    • Base scenario (48-cores/vCPUs) =~ 6 hours
    • Peak scenario (8-cores/vCPUs) =~ 5 hours
    • Peak scenario (48-cores/vCPUs) =~ 8 hours

  • Usage Examples
    The following section provides a few basic examples of how to use the workload profile.

    # Execute the workload profile
    ./VirtualClient --profile=PERF-SPECCPU2026-FPSPEED.json --iterations=1 --logger=csv

PERF-SPECCPU2026-INTRATE.json​

Runs the SPEC CPU 2026 Integer Rate (intrate) benchmark workload on the system focusing on baseline + peak measurements. This workload is an industry standard for evaluating the performance of the CPU for processing calculations.

  • Workload Profile

  • OS/Architecture Platforms

    • linux-x64
    • linux-arm64

  • Dependencies
    The dependencies defined in the 'Dependencies' section of the profile itself are required in order to run the workload operations effectively.

    • Internet connection.

    Additional information on components that exist within the 'Dependencies' section of the profile can be found in the following locations:

  • Profile Parameters
    The following parameters can be optionally supplied on the command line to modify the behaviors of the workload.

    ParameterPurposeDefault value
    CompilerVersionOptional. The version of the compiler to use.13
    IterationCountOptional. The number of iterations for speccpu to run the benchmark/suite.2
    BenchmarksOptional. The benchmark suite or list of benchmarks to run.intrate
    RunPeakOptional. True to run the workload 'Peak' scenario, False to run the workload 'Base' scenario.false (Base)
    ThreadCountOptional. Determines the number of threads to use for running the benchmark.# logical processors
    CopyCountOptional. Determines the number of copies of the benchmark to run concurrently.# logical processors
    BaseOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Base' scenario
    • X64 systems: -g -O2 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O2 -march=armv8.2-a -flto -frecord-gcc-switches
    PeakOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Peak' scenario
    • X64 systems: -g -O3 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O3 -march=armv8.2-a -flto -frecord-gcc-switches

  • Profile Runtimes
    See the 'Metadata' section of the profile for estimated runtimes. These timings represent the length of time required to run a single round of profile actions. These timings can be used to determine minimum required runtimes for the Virtual Client in order to get results. These are often estimates based on the number of system cores. When running the SPEC CPU workload, the timing can be difficult to predict. It is recommended that you avoid using timeouts on the command line and use iterations instead (e.g. --timeout=1440 -> --iterations=1)

    • Base Scenario (8-cores/vCPUs) =~ 10 hours
    • Base scenario (48-cores/vCPUs) =~ 14 hours
    • Peak scenario (8-cores/vCPUs) =~ 18 hours
    • Peak scenario (48-cores/vCPUs) =~ 28 hours

  • Usage Examples
    The following section provides a few basic examples of how to use the workload profile.

    # Execute the workload profile
    ./VirtualClient --profile=PERF-SPECCPU2026-INTRATE.json --iterations=1 --logger=csv

PERF-SPECCPU2026-INTSPEED.json​

Runs the SPEC CPU 2026 Integer Speed (intspeed) benchmark workload on the system focusing on baseline + peak measurements. This workload is an industry standard for evaluating the performance of the CPU for processing calculations.

  • Workload Profile

  • OS/Architecture Platforms

    • linux-x64
    • linux-arm64

  • Dependencies
    The dependencies defined in the 'Dependencies' section of the profile itself are required in order to run the workload operations effectively.

    • Internet connection.

    Additional information on components that exist within the 'Dependencies' section of the profile can be found in the following locations:

  • Profile Parameters
    The following parameters can be optionally supplied on the command line to modify the behaviors of the workload.

    ParameterPurposeDefault value
    CompilerVersionOptional. The version of the compiler to use.13
    IterationCountOptional. The number of iterations for speccpu to run the benchmark/suite.2
    BenchmarksOptional. The benchmark suite or list of benchmarks to run.intspeed
    RunPeakOptional. True to run the workload 'Peak' scenario, False to run the workload 'Base' scenario.false (Base)
    ThreadCountOptional. Determines the number of threads to use for running the benchmark.# logical processors
    CopyCountOptional. Determines the number of copies of the benchmark to run concurrently.# logical processors
    BaseOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Base' scenario
    • X64 systems: -g -O2 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O2 -march=armv8.2-a -flto -frecord-gcc-switches
    PeakOptimizationFlagsOptional. Optimization flags to pass to the GCC compiler when running the 'Peak' scenario
    • X64 systems: -g -O3 -march=x86-64-v3 -flto -frecord-gcc-switches
    • ARM64 systems: -g -O3 -march=armv8.2-a -flto -frecord-gcc-switches

  • Profile Runtimes
    See the 'Metadata' section of the profile for estimated runtimes. These timings represent the length of time required to run a single round of profile actions. These timings can be used to determine minimum required runtimes for the Virtual Client in order to get results. These are often estimates based on the number of system cores. When running the SPEC CPU workload, the timing can be difficult to predict. It is recommended that you avoid using timeouts on the command line and use iterations instead (e.g. --timeout=1440 -> --iterations=1)

    • Base Scenario (8-cores/vCPUs) =~ 4 hours
    • Base scenario (48-cores/vCPUs) =~ 6 hours
    • Peak scenario (8-cores/vCPUs) =~ 5 hours
    • Peak scenario (48-cores/vCPUs) =~ 8 hours

  • Usage Examples
    The following section provides a few basic examples of how to use the workload profile.

    # Execute the workload profile
    ./VirtualClient --profile=PERF-SPECCPU2026-INTSPEED.json --iterations=1 --logger=csv