# Invocation command line: # /home/cpu2017.1.0.2.ic18.0/bin/harness/runcpu --define default-platform-flags --configfile ic18.0-lin-core-avx512-speed-20170901.cfg --define cores=192 --tune base,peak --output_format all --define intspeedaffinity --define drop_caches --nopower --runmode speed --tune base:peak --size refspeed intspeed # output_root was not used for this run ############################################################################ #------------------------------------------------------------------------------ # This is a sample SPEC CPU2017 config file. It is applicable for: # # Compiler name/version: Intel(R) C/C++ and Fortran 18.0.0.128 Compilers for Linux # Operating system version: Redhat Enterprise Linux 7.4, GLIBC 2.17 # Hardware: Intel(R) processors supporting CORE-AVX512 tuning # # If you wish to build your own: # (1) Copy this to a new name # cd %SPEC%/config # copy this.cfg that.cfg # (2) Change items that are marked 'EDIT' (search for it) # # If you have different software or hardware, this config file may not work. # You may find a better config file for your system next to posted results: # http://www.spec.org/cpu2017/results # # Compiler issues: Contact your compiler vendor, not SPEC. # For SPEC help: http://www.spec.org/auto/cpu2017/Docs/techsupport.html #------------------------------------------------------------------------------ #--------- Preprocessor ------------------------------------------------------- # # Optionally edit if you wish: %define build_ncpus 8 # controls number of simultaneous compiles # Used to date the label %define version 20170901 # Used for the optimization tuning part of the label (not required) %if defined(%{noopt}) %define opt_label -noopt %elif defined(%{medopt}) %define opt_label -medopt %else %define opt_label %endif # Used for labeling static linked builds (not required) %if defined(%{static}) %define static_label -static %else %define static_label %endif #--------- Label -------------------------------------------------------------- # Arbitrary string, tags your binaries & directories. # Two Suggestions: # (1) Change this label as you try new ideas. label = ic18.0-lin-core-avx512-speed%{opt_label}%{static_label}-%{version} # (2) Make the label meaningful to YOU. #--------- Global Settings ---------------------------------------------------- # For info, see: # https://www.spec.org/auto/cpu2017/Docs/config.html#fieldname XXX # Example: https://www.spec.org/auto/cpu2017/Docs/config.html#tune ######################################################## # ATTENTION ATTENTION ATTENTION ######################################################## # # NOTE If you change fail_build then PLEASE also # change the line 'define version', because # SPEC review tools use 'label' to track binaries. # ######################################################## # # vvvvvvvvvvvv # do not change unless you read NOTE above fail_build=0 # do not change unless you read NOTE above # ^^^^^^^^^^^^ # do not change unless you read NOTE above # ######################################################## # action = validate command_add_redirect = 1 line_width = 1020 log_line_width = 1020 makeflags = -j%{build_ncpus} output_format = txt,cfg,pdf,csv preenv = 1 tune = base bench_post_setup = sync parallel_test = 1 mean_anyway = 1 reportable = 1 # Set some environment variables preENV_OMP_STACKSIZE = 192M %ifdef %{intspeedaffinity} preENV_KMP_AFFINITY = granularity=fine,scatter %else %if defined(%{smt-on}) preENV_KMP_AFFINITY = granularity=fine,compact,1,0 %else preENV_KMP_AFFINITY = granularity=fine,compact %endif %endif #Reference the flags files flagsurl000=http://www.spec.org/cpu2017/flags/Intel-ic18.0-official-linux64.xml %ifdef %{default-platform-flags} flagsurl001=http://www.spec.org/cpu2017/flags/Lenovo-Platform-SPECcpu2017-Flags-V1.2-SKL-A.xml %endif # Set some environment variables preENV_LD_LIBRARY_PATH = $[top]/lib/ia32:$[top]/lib/intel64:$[top]/je5.0.1-32:$[top]/je5.0.1-64 #--------- Compilers ---------------------------------------------------------- intspeed,fpspeed: CC = icc -m64 -std=c11 CXX = icpc -m64 FC = ifort -m64 # How to say "Show me your version, please" CC_VERSION_OPTION = --version CXX_VERSION_OPTION = --version FC_VERSION_OPTION = --version JEMALLOC32_DIR = /usr/local/je5.0.1-32/lib JEMALLOC64_DIR = /usr/local/je5.0.1-64/lib %if !defined(%{static}) JEMALLOC32 = jemalloc JEMALLOC64 = jemalloc %else JEMALLOC32 = libjemalloc.a JEMALLOC64 = libjemalloc.a %endif #--------- Portability -------------------------------------------------------- intspeed,fpspeed: PORTABILITY = -DSPEC_LP64 600.perlbench_s: CPORTABILITY = -DSPEC_LINUX_X64 621.wrf_s: CPORTABILITY = -DSPEC_CASE_FLAG FPORTABILITY = -convert big_endian 623.xalancbmk_s: CXXPORTABILITY= -DSPEC_LINUX 627.cam4_s: CPORTABILITY= -DSPEC_CASE_FLAG 628.pop2_s: CPORTABILITY = -DSPEC_CASE_FLAG FPORTABILITY = -convert big_endian -assume byterecl #--------- How Many CPUs? ----------------------------------------------------- # Both SPECrate and SPECspeed can test multiple chips / cores / hw threads # - For SPECspeed, you set the number of threads. # See: https://www.spec.org/cpu2017/Docs/system-requirements.html#MultipleCPUs # # q. How many should I set? # a. Unknown, you will have to try it and see! # # To get you started, some suggestions: # # threads - This config file sets a starting point. You could try raising # it. A higher thread count is much more likely to be useful for # fpspeed than for intspeed. # default: %if !defined(%{cores}) %error please add --define cores=[ncores] to runcpu invocation %endif threads = %{cores} # EDIT to change number of OpenMP threads (see above) %ifdef %{smt-on} %define numpeakthreads %{cores} * 2 %endif intspeed: %define numxzthreads %{cores} * 2 %ifdef %{smt-on} threads = %{numxzthreads} %else threads = %{cores} %endif #-------- Tuning Flags ---------------------------------------------- intrate,fprate: #reconsider using a rate config file instead of a speed config file fail=1 intspeed,fpspeed: # Compiler tuning toggle, corresponds with isa_ext preprocessor cfg var SSE = -xCORE-AVX512 EXTRA_FOPTIMIZE = -nostandard-realloc-lhs -align array32byte %if defined(%{noopt}) OPT_ROOT = -O0 %elif defined(%{medopt}) OPT_ROOT = -O2 %else OPT_ROOT = -ipo -O3 -no-prec-div %endif FAST_NO_STATIC = $(SSE) $(OPT_ROOT) FORT_FAST_NO_STATIC = $(SSE) $(OPT_ROOT) FAST = $(SSE) $(OPT_ROOT) -static FORT_FAST = $(SSE) $(OPT_ROOT) -static intspeed=base: EXTRA_LIBS = -L$(JEMALLOC64_DIR) -l$(JEMALLOC64) EXTRA_LDFLAGS = -Wl,-z,muldefs intspeed=base=default: %if !defined(%{static}) COPTIMIZE = $(FAST_NO_STATIC) -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP CXXOPTIMIZE = $(FAST_NO_STATIC) -qopt-mem-layout-trans=3 FOPTIMIZE = $(FORT_FAST_NO_STATIC) -qopt-mem-layout-trans=3 %if defined(%{noopt}) COPTIMIZE = $(OPT_ROOT) -qopenmp -DSPEC_OPENMP CXXOPTIMIZE = $(OPT_ROOT) FOPTIMIZE = $(OPT_ROOT) %endif %else COPTIMIZE = $(FAST) -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP CXXOPTIMIZE = $(FAST) -qopt-mem-layout-trans=3 FOPTIMIZE = $(FORT_FAST) -qopt-mem-layout-trans=3 %if defined(%{noopt}) COPTIMIZE = $(OPT_ROOT) -qopenmp -DSPEC_OPENMP -static CXXOPTIMIZE = $(OPT_ROOT) -static FOPTIMIZE = $(OPT_ROOT) -static %endif %endif fpspeed=base=default: %if !defined(%{static}) COPTIMIZE = $(FAST_NO_STATIC) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP CXXOPTIMIZE = $(FAST_NO_STATIC) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP FOPTIMIZE = $(FORT_FAST_NO_STATIC) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP %if defined(%{noopt}) COPTIMIZE = $(OPT_ROOT) -qopenmp -DSPEC_OPENMP CXXOPTIMIZE = $(OPT_ROOT) -qopenmp -DSPEC_OPENMP FOPTIMIZE = $(OPT_ROOT) -qopenmp -DSPEC_OPENMP %endif %else COPTIMIZE = $(FAST) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP CXXOPTIMIZE = $(FAST) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP FOPTIMIZE = $(FORT_FAST) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP %if defined(%{noopt}) COPTIMIZE = $(OPT_ROOT) -qopenmp -DSPEC_OPENMP -static CXXOPTIMIZE = $(OPT_ROOT) -qopenmp -DSPEC_OPENMP -static FOPTIMIZE = $(OPT_ROOT) -qopenmp -DSPEC_OPENMP -static %endif %endif default=peak=default: %if defined(%{noopt}) || defined(%{medopt}) #not building or executing peak against less opt fail=1 %endif PASS1_CFLAGS = -prof-gen PASS2_CFLAGS = -prof-use PASS1_FFLAGS = -prof-gen PASS2_FFLAGS = -prof-use PASS1_CXXFLAGS = -prof-gen PASS2_CXXFLAGS = -prof-use PASS1_LDFLAGS = -prof-gen PASS2_LDFLAGS = -prof-use OPT_ROOT = -ipo -O3 -no-prec-div FAST_NO_STATIC_PASS1 = $(OPT_ROOT) FORT_FAST_NO_STATIC_PASS1 = $(OPT_ROOT) FAST_PASS1 = $(OPT_ROOT) -static FORT_FAST_PASS1 = $(OPT_ROOT) -static FAST_NO_STATIC_PASS2 = $(SSE) $(OPT_ROOT) FORT_FAST_NO_STATIC_PASS2 = $(SSE) $(OPT_ROOT) FAST_PASS2 = $(SSE) $(OPT_ROOT) -static FORT_FAST_PASS2 = $(SSE) $(OPT_ROOT) -static fpspeed,intspeed_any_c=peak: OPT_ROOT_PASS1 = -O2 OPT_ROOT_PASS2 = -ipo -O3 -no-prec-div FAST_NO_STATIC_PASS1 = $(OPT_ROOT_PASS1) FORT_FAST_NO_STATIC_PASS1 = $(OPT_ROOT_PASS1) FAST_PASS1 = $(OPT_ROOT_PASS1) -static FORT_FAST_PASS1 = $(OPT_ROOT_PASS1) -static FAST_NO_STATIC_PASS2 = $(SSE) $(OPT_ROOT_PASS2) FORT_FAST_NO_STATIC_PASS2 = $(SSE) $(OPT_ROOT_PASS2) FAST_PASS2 = $(SSE) $(OPT_ROOT_PASS2) -static FORT_FAST_PASS2 = $(SSE) $(OPT_ROOT_PASS2) -static intspeed=peak: EXTRA_LIBS = -L$(JEMALLOC64_DIR) -l$(JEMALLOC64) EXTRA_LDFLAGS = -Wl,-z,muldefs 623.xalancbmk_s=peak: ENV_OMP_STACKSIZE = 128M CC = icc -m32 -std=c11 -L/opt/intel/compilers_and_libraries_2018/linux/lib/ia32 CXX = icpc -m32 -L/opt/intel/compilers_and_libraries_2018/linux/lib/ia32 FC = ifort -m32 -L/opt/intel/compilers_and_libraries_2018/linux/lib/ia32 EXTRA_LIBS = -L$(JEMALLOC32_DIR) -l$(JEMALLOC32) EXTRA_LDFLAGS = -Wl,-z,muldefs #Change portability for 32b, large file support for safety PORTABILITY = -D_FILE_OFFSET_BITS=64 605.mcf_s,631.deepsjeng_s=peak: FAST_NO_STATIC_PASS1 = $(OPT_ROOT) FORT_FAST_NO_STATIC_PASS1 = $(OPT_ROOT) FAST_PASS1 = $(OPT_ROOT) -static FORT_FAST_PASS1 = $(OPT_ROOT) -static FAST_NO_STATIC_PASS2 = $(SSE) $(OPT_ROOT) FORT_FAST_NO_STATIC_PASS2 = $(SSE) $(OPT_ROOT) FAST_PASS2 = $(SSE) $(OPT_ROOT) -static FORT_FAST_PASS2 = $(SSE) $(OPT_ROOT) -static intspeed=peak=default: %if !defined(%{static}) PASS1_COPTIMIZE = $(FAST_NO_STATIC_PASS1) -qopt-mem-layout-trans=3 -DSPEC_SUPPRESS_OPENMP PASS1_CXXOPTIMIZE = $(FAST_NO_STATIC_PASS1) -qopt-mem-layout-trans=3 -DSPEC_SUPPRESS_OPENMP PASS1_FOPTIMIZE = $(FORT_FAST_NO_STATIC_PASS1) -qopt-mem-layout-trans=3 -DSPEC_SUPPRESS_OPENMP PASS2_COPTIMIZE = $(FAST_NO_STATIC_PASS2) -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP PASS2_CXXOPTIMIZE = $(FAST_NO_STATIC_PASS2) -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP PASS2_FOPTIMIZE = $(FORT_FAST_NO_STATIC_PASS2) -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP %else PASS1_COPTIMIZE = $(FAST_PASS1) -qopt-mem-layout-trans=3 -DSPEC_SUPPRESS_OPENMP PASS1_CXXOPTIMIZE = $(FAST_PASS1) -qopt-mem-layout-trans=3 -DSPEC_SUPPRESS_OPENMP PASS1_FOPTIMIZE = $(FORT_FAST_PASS1) -qopt-mem-layout-trans=3 -DSPEC_SUPPRESS_OPENMP PASS2_COPTIMIZE = $(FAST_PASS2) -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP PASS2_CXXOPTIMIZE = $(FAST_PASS2) -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP PASS2_FOPTIMIZE = $(FORT_FAST_PASS2) -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP %endif fpspeed=peak=default: %if !defined(%{static}) PASS1_COPTIMIZE = $(FAST_NO_STATIC_PASS1) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -DSPEC_SUPPRESS_OPENMP PASS1_CXXOPTIMIZE = $(FAST_NO_STATIC_PASS1) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -DSPEC_SUPPRESS_OPENMP PASS1_FOPTIMIZE = $(FORT_FAST_NO_STATIC_PASS1) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -DSPEC_SUPPRESS_OPENMP PASS2_COPTIMIZE = $(FAST_NO_STATIC_PASS2) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP PASS2_CXXOPTIMIZE = $(FAST_NO_STATIC_PASS2) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP PASS2_FOPTIMIZE = $(FORT_FAST_NO_STATIC_PASS2) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP %else PASS1_COPTIMIZE = $(FAST_PASS1) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -DSPEC_SUPPRESS_OPENMP PASS1_CXXOPTIMIZE = $(FAST_PASS1) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -DSPEC_SUPPRESS_OPENMP PASS1_FOPTIMIZE = $(FORT_FAST_PASS1) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -DSPEC_SUPPRESS_OPENMP PASS2_COPTIMIZE = $(FAST_PASS2) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP PASS2_CXXOPTIMIZE = $(FAST_PASS2) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP PASS2_FOPTIMIZE = $(FORT_FAST_PASS2) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP %endif #some benchmarks do not mind sharing cores %ifdef %{smt-on} 627.cam4_s,628.pop2_s,644.nab_s=peak: threads = %{numpeakthreads} %endif 600.perlbench_s=peak=default: #[56]00.perlbench[rs] has non standard signed overflow EXTRA_OPTIMIZE = -fno-strict-overflow #fp no FDO 627.cam4_s,638.imagick_s,644.nab_s=peak: feedback = 0 %if !defined(%{static}) COPTIMIZE = $(FAST_NO_STATIC) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP CXXOPTIMIZE = $(FAST_NO_STATIC) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP FOPTIMIZE = $(FORT_FAST_NO_STATIC) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP %else COPTIMIZE = $(FAST) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP CXXOPTIMIZE = $(FAST) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP FOPTIMIZE = $(FORT_FAST) -qopt-prefetch -ffinite-math-only -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP %endif #int no FDO 625.x264_s,648.exchange2_s=peak: feedback = 0 %if !defined(%{static}) COPTIMIZE = $(FAST_NO_STATIC) -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP CXXOPTIMIZE = $(FAST_NO_STATIC) -qopt-mem-layout-trans=3 FOPTIMIZE = $(FORT_FAST_NO_STATIC) -qopt-mem-layout-trans=3 %else COPTIMIZE = $(FAST) -qopt-mem-layout-trans=3 -qopenmp -DSPEC_OPENMP CXXOPTIMIZE = $(FAST) -qopt-mem-layout-trans=3 FOPTIMIZE = $(FORT_FAST) -qopt-mem-layout-trans=3 %endif #------------------------------------------------------------------------------- # Tester and System Descriptions - EDIT the sections below #------------------------------------------------------------------------------- # For info about any field, see # https://www.spec.org/auto/cpu2017/Docs/config.html#fieldname # Example: https://www.spec.org/auto/cpu2017/Docs/config.html#hw_memory #------------------------------------------------------------------------------- #--------- If you install new compilers, EDIT this section -------------------- intrate,intspeed,fpspeed,fprate: sw_compiler000 = C/C++: Version 18.0.0.128 of Intel C/C++ sw_compiler001 = Compiler for Linux; sw_compiler002 = Fortran: Version 18.0.0.128 of Intel Fortran sw_compiler003 = Compiler for Linux sw_base_ptrsize = 64-bit fprate,fpspeed: sw_peak_ptrsize = 64-bit intrate,intspeed: sw_peak_ptrsize = 32/64-bit sw_other000 = jemalloc: jemalloc memory allocator library sw_other001 = V5.0.1; sw_other002 = jemalloc: configured and built at default for sw_other003 = 32bit (i686) and 64bit (x86_64) targets; sw_other004 = jemalloc: built with the RedHat Enterprise 7.4, sw_other005 = and the system compiler gcc 4.8.5; sw_other006 = jemalloc: sources avilable from jemalloc.net or sw_other007 = https://github.com/jemalloc/jemalloc/releases #--------- EDIT Your info --------------------------------------------------------- # To understand the difference between hw_vendor/sponsor/tester, see: # www.spec.org/auto/cpu2017/Docs/config.html#test_sponsor intrate,intspeed,fprate,fpspeed: # Important: keep this line hw_vendor = Lenovo Global Technology tester = Lenovo Global Technology test_sponsor = Lenovo Global Technology license_num = 9017 prepared_by = Lenovo Global Technology #--------- EDIT system availability dates and system information ---------------------------------------------- intrate,intspeed,fprate,fpspeed: # Important: keep this line # Example # Brief info about field hw_avail = Sep-2017 sw_avail = Sep-2017 hw_cpu_nominal_mhz = 2700 hw_cpu_max_mhz = 3700 hw_ncores = 192 hw_nthreadspercore = 1 hw_ncpuorder = 2,4,8 chips hw_model000 = ThinkSystem SR950 hw_model001 = (2.70 GHz, Intel Xeon Platinum 8168) hw_other = None # sw_other = # TurboHeap Library V8.1 # Other perf-relevant sw, or "None" hw_pcache = 32 KB I + 32 KB D on chip per core hw_scache = 1 MB I+D on chip per core hw_tcache = 33 MB I+D on chip per chip hw_ocache = None # hw_memory001 = # 4 TB (256 x 16 GB 2Rx4 PC4-2133P-R, # N GB (M x N GB nRxn # hw_memory002 = # running at 1600 MHz) # PCn-nnnnnR-n[, ECC and other info]) #--------- Sysinfo fields - You may need to adjust this section --------------- # Note: Some commented-out fields above are automatically set to preliminary # values by sysinfo # www.spec.org/auto/cpu2017/Docs/config.html#sysinfo # Uncomment lines for which you already have a better answer than sysinfo # intrate,intspeed,fprate,fpspeed: # Important: keep this line # Example # Brief info about field # hw_cpu_name = # Intel Xeon E9-9999 v9 # chip name # hw_disk = # 9 x 9 TB SATA III 9999 RPM # Size, type, other perf-relevant info # hw_nchips = # 99 # number chips enabled # sw_file = # ext99 # File system # sw_state = # Run level 99 # Software state. # sw_os001 = # Linux Sailboat # Operating system # sw_os002 = # Distribution 7.2 SP1 # and version #--------- EDIT Intel Recommended Fields - You may wish to adjust this section ----- default: notes_025 = Binaries compiled on a system with 1x Intel Core i7-4790 CPU + 32GB RAM notes_030 = memory using Redhat Enterprise Linux 7.4 notes_os_000 = Stack size set to unlimited using "ulimit -s unlimited" #Ease of tagging results from the runcpu command line with commands run outside of runcpu harneess #Adjust as necessary for your SUT %if defined(%{THP_enabled}) notes_200 = Transparent Huge Pages enabled with: notes_201 = echo always > /sys/kernel/mm/transparent_hugepage/enabled %elif defined(%{THP_disabled}) notes_200 = Transparent Huge Pages disabled with: notes_201 = echo never > /sys/kernel/mm/transparent_hugepage/enabled %else notes_035 = Transparent Huge Pages enabled by default %endif %ifdef %{drop_caches} notes_040 = Prior to runcpu invocation notes_045 = Filesystem page cache synced and cleared with: notes_050 = sync; echo 3> /proc/sys/vm/drop_caches %endif %ifdef %{invoke_with_interleave} notes_205 = runcpu command invoked through numactl i.e.: notes_206 = numactl --interleave=all runcpu %endif intrate,fprate: %if defined(%{no-numa}) notes_submit_000 = The taskset mechanism was used to bind copies to processors. The config file option 'submit' notes_submit_001 = was used to generate taskset commands to bind each copy to a specific processor. notes_submit_002 = For details, please see the config file. %else notes_submit_000 = The numactl mechanism was used to bind copies to processors. The config file option 'submit' notes_submit_001 = was used to generate numactl commands to bind each copy to a specific processor. notes_submit_002 = For details, please see the config file. %endif # The following settings were obtained by running the sysinfo_program # 'specperl $[top]/bin/sysinfo' (sysinfo:SHA:ecd2bef08f316af97f5a7768b641e2a3307c1b4b68efb5a57fa76367d790d233) default: notes_plat_sysinfo_000 = Sysinfo program /home/cpu2017.1.0.2.ic18.0/bin/sysinfo notes_plat_sysinfo_005 = Rev: r5797 of 2017-06-14 96c45e4568ad54c135fd618bcc091c0f notes_plat_sysinfo_010 = running on linux-boxi Mon Dec 11 16:06:49 2017 notes_plat_sysinfo_015 = notes_plat_sysinfo_020 = SUT (System Under Test) info as seen by some common utilities. notes_plat_sysinfo_025 = For more information on this section, see notes_plat_sysinfo_030 = https://www.spec.org/cpu2017/Docs/config.html#sysinfo notes_plat_sysinfo_035 = notes_plat_sysinfo_040 = From /proc/cpuinfo notes_plat_sysinfo_045 = model name : Intel(R) Xeon(R) Platinum 8168 CPU @ 2.70GHz notes_plat_sysinfo_050 = 8 "physical id"s (chips) notes_plat_sysinfo_055 = 192 "processors" notes_plat_sysinfo_060 = cores, siblings (Caution: counting these is hw and system dependent. The following notes_plat_sysinfo_065 = excerpts from /proc/cpuinfo might not be reliable. Use with caution.) notes_plat_sysinfo_070 = cpu cores : 24 notes_plat_sysinfo_075 = siblings : 24 notes_plat_sysinfo_080 = physical 0: cores 0 1 2 3 4 5 8 9 10 11 12 13 16 17 18 19 20 21 24 25 26 27 28 29 notes_plat_sysinfo_085 = physical 1: cores 0 1 2 3 4 5 8 9 10 11 12 13 16 17 18 19 20 21 24 25 26 27 28 29 notes_plat_sysinfo_090 = physical 2: cores 0 1 2 3 4 5 8 9 10 11 12 13 16 17 18 19 20 21 24 25 26 27 28 29 notes_plat_sysinfo_095 = physical 3: cores 0 1 2 3 4 5 8 9 10 11 12 13 16 17 18 19 20 21 24 25 26 27 28 29 notes_plat_sysinfo_100 = physical 4: cores 0 1 2 3 4 5 8 9 10 11 12 13 16 17 18 19 20 21 24 25 26 27 28 29 notes_plat_sysinfo_105 = physical 5: cores 0 1 2 3 4 5 8 9 10 11 12 13 16 17 18 19 20 21 24 25 26 27 28 29 notes_plat_sysinfo_110 = physical 6: cores 0 1 2 3 4 5 8 9 10 11 12 13 16 17 18 19 20 21 24 25 26 27 28 29 notes_plat_sysinfo_115 = physical 7: cores 0 1 2 3 4 5 8 9 10 11 12 13 16 17 18 19 20 21 24 25 26 27 28 29 notes_plat_sysinfo_120 = notes_plat_sysinfo_125 = From lscpu: notes_plat_sysinfo_130 = Architecture: x86_64 notes_plat_sysinfo_135 = CPU op-mode(s): 32-bit, 64-bit notes_plat_sysinfo_140 = Byte Order: Little Endian notes_plat_sysinfo_145 = CPU(s): 192 notes_plat_sysinfo_150 = On-line CPU(s) list: 0-191 notes_plat_sysinfo_155 = Thread(s) per core: 1 notes_plat_sysinfo_160 = Core(s) per socket: 24 notes_plat_sysinfo_165 = Socket(s): 8 notes_plat_sysinfo_170 = NUMA node(s): 8 notes_plat_sysinfo_175 = Vendor ID: GenuineIntel notes_plat_sysinfo_180 = CPU family: 6 notes_plat_sysinfo_185 = Model: 85 notes_plat_sysinfo_190 = Model name: Intel(R) Xeon(R) Platinum 8168 CPU @ 2.70GHz notes_plat_sysinfo_195 = Stepping: 4 notes_plat_sysinfo_200 = CPU MHz: 2693.676 notes_plat_sysinfo_205 = BogoMIPS: 5387.35 notes_plat_sysinfo_210 = Virtualization: VT-x notes_plat_sysinfo_215 = L1d cache: 32K notes_plat_sysinfo_220 = L1i cache: 32K notes_plat_sysinfo_225 = L2 cache: 1024K notes_plat_sysinfo_230 = L3 cache: 33792K notes_plat_sysinfo_235 = NUMA node0 CPU(s): 0-23 notes_plat_sysinfo_240 = NUMA node1 CPU(s): 24-47 notes_plat_sysinfo_245 = NUMA node2 CPU(s): 48-71 notes_plat_sysinfo_250 = NUMA node3 CPU(s): 72-95 notes_plat_sysinfo_255 = NUMA node4 CPU(s): 96-119 notes_plat_sysinfo_260 = NUMA node5 CPU(s): 120-143 notes_plat_sysinfo_265 = NUMA node6 CPU(s): 144-167 notes_plat_sysinfo_270 = NUMA node7 CPU(s): 168-191 notes_plat_sysinfo_275 = Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov notes_plat_sysinfo_280 = pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp notes_plat_sysinfo_285 = lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc notes_plat_sysinfo_290 = aperfmperf eagerfpu pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 sdbg notes_plat_sysinfo_295 = fma cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes notes_plat_sysinfo_300 = xsave avx f16c rdrand lahf_lm abm 3dnowprefetch ida arat epb pln pts dtherm intel_pt notes_plat_sysinfo_305 = tpr_shadow vnmi flexpriority ept vpid fsgsbase tsc_adjust bmi1 hle avx2 smep bmi2 notes_plat_sysinfo_310 = erms invpcid rtm cqm mpx avx512f avx512dq rdseed adx smap clflushopt clwb avx512cd notes_plat_sysinfo_315 = avx512bw avx512vl xsaveopt xsavec xgetbv1 cqm_llc cqm_occup_llc notes_plat_sysinfo_320 = notes_plat_sysinfo_325 = /proc/cpuinfo cache data notes_plat_sysinfo_330 = cache size : 33792 KB notes_plat_sysinfo_335 = notes_plat_sysinfo_340 = From numactl --hardware WARNING: a numactl 'node' might or might not correspond to a notes_plat_sysinfo_345 = physical chip. notes_plat_sysinfo_350 = available: 8 nodes (0-7) notes_plat_sysinfo_355 = node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 notes_plat_sysinfo_360 = node 0 size: 386513 MB notes_plat_sysinfo_365 = node 0 free: 385491 MB notes_plat_sysinfo_370 = node 1 cpus: 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 notes_plat_sysinfo_375 = node 1 size: 387057 MB notes_plat_sysinfo_380 = node 1 free: 386102 MB notes_plat_sysinfo_385 = node 2 cpus: 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 notes_plat_sysinfo_390 = node 2 size: 387057 MB notes_plat_sysinfo_395 = node 2 free: 386118 MB notes_plat_sysinfo_400 = node 3 cpus: 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 notes_plat_sysinfo_405 = node 3 size: 387057 MB notes_plat_sysinfo_410 = node 3 free: 386004 MB notes_plat_sysinfo_415 = node 4 cpus: 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 notes_plat_sysinfo_420 = 115 116 117 118 119 notes_plat_sysinfo_425 = node 4 size: 387057 MB notes_plat_sysinfo_430 = node 4 free: 386160 MB notes_plat_sysinfo_435 = node 5 cpus: 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 notes_plat_sysinfo_440 = 138 139 140 141 142 143 notes_plat_sysinfo_445 = node 5 size: 387057 MB notes_plat_sysinfo_450 = node 5 free: 385880 MB notes_plat_sysinfo_455 = node 6 cpus: 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 notes_plat_sysinfo_460 = 162 163 164 165 166 167 notes_plat_sysinfo_465 = node 6 size: 387057 MB notes_plat_sysinfo_470 = node 6 free: 386139 MB notes_plat_sysinfo_475 = node 7 cpus: 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 notes_plat_sysinfo_480 = 186 187 188 189 190 191 notes_plat_sysinfo_485 = node 7 size: 387052 MB notes_plat_sysinfo_490 = node 7 free: 386052 MB notes_plat_sysinfo_495 = node distances: notes_plat_sysinfo_500 = node 0 1 2 3 4 5 6 7 notes_plat_sysinfo_505 = 0: 10 21 31 21 21 31 31 31 notes_plat_sysinfo_510 = 1: 21 10 21 31 31 21 31 31 notes_plat_sysinfo_515 = 2: 31 21 10 21 31 31 21 31 notes_plat_sysinfo_520 = 3: 21 31 21 10 31 31 31 21 notes_plat_sysinfo_525 = 4: 21 31 31 31 10 21 21 31 notes_plat_sysinfo_530 = 5: 31 21 31 31 21 10 31 21 notes_plat_sysinfo_535 = 6: 31 31 21 31 21 31 10 21 notes_plat_sysinfo_540 = 7: 31 31 31 21 31 21 21 10 notes_plat_sysinfo_545 = notes_plat_sysinfo_550 = From /proc/meminfo notes_plat_sysinfo_555 = MemTotal: 3170210080 kB notes_plat_sysinfo_560 = HugePages_Total: 0 notes_plat_sysinfo_565 = Hugepagesize: 2048 kB notes_plat_sysinfo_570 = notes_plat_sysinfo_575 = From /etc/*release* /etc/*version* notes_plat_sysinfo_580 = SuSE-release: notes_plat_sysinfo_585 = SUSE Linux Enterprise Server 12 (x86_64) notes_plat_sysinfo_590 = VERSION = 12 notes_plat_sysinfo_595 = PATCHLEVEL = 2 notes_plat_sysinfo_600 = # This file is deprecated and will be removed in a future service pack or release. notes_plat_sysinfo_605 = # Please check /etc/os-release for details about this release. notes_plat_sysinfo_610 = os-release: notes_plat_sysinfo_615 = NAME="SLES" notes_plat_sysinfo_620 = VERSION="12-SP2" notes_plat_sysinfo_625 = VERSION_ID="12.2" notes_plat_sysinfo_630 = PRETTY_NAME="SUSE Linux Enterprise Server 12 SP2" notes_plat_sysinfo_635 = ID="sles" notes_plat_sysinfo_640 = ANSI_COLOR="0;32" notes_plat_sysinfo_645 = CPE_NAME="cpe:/o:suse:sles:12:sp2" notes_plat_sysinfo_650 = notes_plat_sysinfo_655 = uname -a: notes_plat_sysinfo_660 = Linux linux-boxi 4.4.21-69-default #1 SMP Tue Oct 25 10:58:20 UTC 2016 (9464f67) notes_plat_sysinfo_665 = x86_64 x86_64 x86_64 GNU/Linux notes_plat_sysinfo_670 = notes_plat_sysinfo_675 = run-level 3 Dec 11 16:05 notes_plat_sysinfo_680 = notes_plat_sysinfo_685 = SPEC is set to: /home/cpu2017.1.0.2.ic18.0 notes_plat_sysinfo_690 = Filesystem Type Size Used Avail Use% Mounted on notes_plat_sysinfo_695 = /dev/sdb2 btrfs 744G 25G 719G 4% /home notes_plat_sysinfo_700 = notes_plat_sysinfo_705 = Additional information from dmidecode follows. WARNING: Use caution when you interpret notes_plat_sysinfo_710 = this section. The 'dmidecode' program reads system data which is "intended to allow notes_plat_sysinfo_715 = hardware to be accurately determined", but the intent may not be met, as there are notes_plat_sysinfo_720 = frequent changes to hardware, firmware, and the "DMTF SMBIOS" standard. notes_plat_sysinfo_725 = BIOS Lenovo -[PSE105X-1.00]- 08/17/2017 notes_plat_sysinfo_730 = Memory: notes_plat_sysinfo_735 = 96x Samsung M393A4K40BB2-CTD 32 GB 2 rank 2666 notes_plat_sysinfo_740 = notes_plat_sysinfo_745 = (End of data from sysinfo program) hw_cpu_name = Intel Xeon Platinum 8168 hw_disk = 1 x 800 GB SAS SSD hw_memory = 3 TB (96 x 32 GB 2Rx4 PC4-2666V-R) hw_nchips = 8 prepared_by = root (is never output, only tags rawfile) fw_bios = Lenovo BIOS Version PSE105X 1.00 released Aug-2017 sw_file = btrfs sw_os000 = SUSE Linux Enterprise Server 12 SP2 (x86_64) sw_os001 = Kernel 4.4.21-69-default sw_state = Run level 3 (multi-user) # End of settings added by sysinfo_program # The following section was added automatically, and contains settings that # did not appear in the original configuration file, but were added to the # raw file after the run. default: notes_000 =Environment variables set by runcpu before the start of the run: notes_005 =KMP_AFFINITY = "granularity=fine,scatter" notes_010 =LD_LIBRARY_PATH = "/home/cpu2017.1.0.2.ic18.0/lib/ia32:/home/cpu2017.1.0.2.ic18.0/lib/intel64" notes_015 =LD_LIBRARY_PATH = "$LD_LIBRARY_PATH:/home/cpu2017.1.0.2.ic18.0/je5.0.1-32:/home/cpu2017.1.0.2.ic18.0/je5.0.1-64" notes_020 = notes_plat_form_000 =BIOS configuration: notes_plat_form_005 =Choose Operating Mode set to Maximum Performance notes_plat_form_010 =Hyper-Threading set to Disable notes_plat_form_015 =MONITORMWAIT set to Enable notes_plat_form_020 =Execute Disable Bit set to Disable notes_plat_form_025 =Intel Virtualization Technology set to Disable notes_plat_form_030 =Uncore Frequency Scaling set to Disable notes_plat_form_035 =XPT Prefetcher set to Enable notes_plat_form_040 =UPI Prefetcher set to Disable notes_plat_form_045 =Stale AtoS set to Enable