# Invocation command line: # /local-home/omp2001/bin/runspec --reportable -a validate -n 3 -o asc -I -i ref -T base,peak -c sgi-uv2-12.1-tune.cfg large ############################################################################ ############################################################################ VENDOR = SGI action = validate ext = hfeng_intel.12.1.tune PATHSEP = / check_md5 = 1 env_vars = 1 iterations = 3 level = 40 output_format = config,asc backup_config = 0 mean_anyway = 1 # # These are listed as benchmark-tuning-extension-machine # default=default=default=default: CC = icc FC = ifort OBJ = .o # # portability & libraries # default=default=default=default: ENV_KMP_LIBRARY=turnaround ENV_KMP_BLOCKTIME=infinite ENV_KMP_STACKSIZE=31M ENV_OMP_DYNAMIC=FALSE ONESTEP=yes 318.galgel_m=default=default=default: EXTRA_FFLAGS = -FI -132 notes061 = 318.galgel_m: -FI -132 330.art_m=default=default=default: EXTRA_CFLAGS = -DINTS_PER_CACHELINE=16 -DDBLS_PER_CACHELINE=8 notes066 = 330.art_m: -DINTS_PER_CACHELINE=16 -DDBLS_PER_CACHELINE=8 331.art_l=default=default=default: EXTRA_CFLAGS = -DINTS_PER_CACHELINE=16 -DDBLS_PER_CACHELINE=8 notes067 = 331.art_l: -DINTS_PER_CACHELINE=16 -DDBLS_PER_CACHELINE=8 default=default=default=default: notes030 = ======================================================================== notes040 = BIOS settings notes: notes042 = Intel Turbo Boost Technology (Turbo) : Enabled (Max 2.80GHz) notes045 = notes100 = ======================================================================== notes101 = General Notes and Enviroment variables notes105 = export KMP_LIBRARY=turnaround notes106 = export KMP_STACKSIZE=31M notes107 = export KMP_BLOCKTIME=infinite notes108 = export OMP_DYNAMIC=FALSE notes109 = ONESTEP=yes notes110 = ulimit -s unlimited notes111 = notes115 = For compiler/openmp flags description please refer: notes116 = SGI-20100816-Linux-Intel11.1-intel64.html notes117 = notes130 = Base optimization flags and environment variables: notes230 = Peak optimization flags and environment variables: notes300 = Peak per-benchmark optimization flags and environment variables: ################################################################ # Base tuning default optimization # # # ################################################################ medium=base=default=default: OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp COPTIMIZE=-ansi-alias #ENV_OMP_NUM_THREADS=32 ENV_KMP_AFFINITY=compact,0 notes150 = ======================================================================== notes151 = Medium: notes155 = OPTIMIZE = -O3 -xSSE4.2 -ipo1 -openmp notes156 = COPTIMIZE = -ansi-alias notes157 = export OMP_NUM_THREADS=32 notes165 = export KMP_AFFINITY=compact,1 notes166 = large=base=default=default: OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp -mcmodel=medium -shared-intel COPTIMIZE=-ansi-alias ENV_KMP_AFFINITY=scatter,1 notes170 = ======================================================================== notes171 = Large: notes175 = OPTIMIZE = -O3 -xSSE4.2 -ipo1 -openmp -mcmodel=medium -shared-intel notes176 = COPTIMIZE = -ansi-alias notes190 = export KMP_AFFINITY=scatter,1 notes191 = ################################################################ # Peak tuning default optimization # # # ################################################################ medium=peak=default=default: OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp -rcd ENV_KMP_AFFINITY=compact,1 notes250 = ======================================================================== notes251 = Medium: notes255 = OPTIMIZE = -O3 -xSSE4.2 -ipo1 -openmp -rcd notes265 = export KMP_AFFINITY=compact,1 notes266 = large=peak=default=default: OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp -rcd ENV_KMP_AFFINITY=scatter,1 notes270 = ======================================================================== notes271 = Large: notes275 = OPTIMIZE = -O3 -xSSE4.2 -ipo1 -openmp -rcd notes285 = export KMP_AFFINITY=compact,1 notes291 = ################################################################ # Peak tuning per-benchmark optimization # # # ################################################################ 310.wupwise_m=peak=default=default: ENV_OMP_NUM_THREADS=32 notes310_0 = ======================================================================== notes310_1 = 310.wupwise_m notes310_2 = export OMP_NUM_THREADS=32 notes310_3 = 311.wupwise_l=peak=default=default: basepeak=yes notes311_0 = ======================================================================== notes311_1 = 311.wupwise_l notes311_2 = basepeak=1 notes311_4 = 312.swim_m=peak=default=default: OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp -opt-streaming-stores always -align -rcd srcalt = ompl.32 ENV_OMP_NUM_THREADS=32 notes312_0 = ======================================================================== notes312_1 = 312.swim_m notes312_2 = OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp -opt-streaming-stores always -align -rcd notes312_3 = srcalt = ompl.32 notes312_4 = export OMP_NUM_THREADS=32 notes312_5 = 313.swim_l=peak=default=default: basepeak=1 notes313_0 = ======================================================================== notes313_1 = 313.swim_m notes313_2 = basepeak=1 notes313_5 = 314.mgrid_m=peak=default=default: OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp -fno-alias -rcd ENV_OMP_NUM_THREADS=32 notes314_0 = ======================================================================== notes314_1 = 314.mgrid_m notes314_2 = OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp -fno-alias -rcd notes314_3 = export OMP_NUM_THREADS=32 notes314_4 = 315.mgrid_l=peak=default=default: OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp -rcd ENV_OMP_NUM_THREADS=256 notes315_0 = ======================================================================== notes315_1 = 315.mgrid_l notes315_2 = OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp -rcd notes315_4 = export OMP_NUM_THREADS=256 notes314_5 = 316.applu_m=peak=default=default: ENV_KMP_AFFINITY=scatter,1 notes316_0 = ======================================================================== notes316_1 = 316.applu_m notes316_2 = export KMP_AFFINITY=scatter,1 notes316_3 = 317.applu_l=peak=default=default: OPTIMIZE=-O3 -xAVX -ipo1 -openmp -rcd -mcmodel=medium -shared-intel ENV_OMP_NUM_THREADS=416 notes317_0 = ======================================================================== notes317_1 = 317.applu_l notes317_2 = OPTIMIZE=-O3 -xAVX -ipo1 -openmp -rcd -mcmodel=medium -shared-intel notes317_4 = export OMP_NUM_THREADS=416 notes317_5 = 318.galgel_m=peak=default=default: ENV_OMP_NUM_THREADS=16 ENV_KMP_LIBRARY=turnaround ENV_KMP_AFFINITY=granularity=fine,scatter,0 RM_SOURCES=lapak.f90 EXTRA_LIBS=-lmkl_intel_lp64 -lmkl_intel_thread -lmkl_core notes318_0 = ======================================================================== notes318_1 = 318.galgel_m notes318_2 = export OMP_NUM_THREADS=32 notes318_4 = 320.equake_m=peak=default=default: ENV_OMP_NUM_THREADS=32 notes320_0 = ======================================================================== notes320_1 = 320.equake_m notes320_3 = export OMP_NUM_THREADS=32 notes320_4 = 321.equake_l=peak=default=default: OPTIMIZE=-O3 -xAVX -ipo1 -openmp -mcmodel=medium -shared-intel ENV_OMP_NUM_THREADS=256 notes321_0 = ======================================================================== notes321_1 = 321.equake_l notes321_2 = OPTIMIZE=-O3 -xAVX -ipo1 -openmp -mcmodel=medium -shared-intel notes321_3 = export OMP_NUM_THREADS=256 notes321_4 = 324.apsi_m=peak=default=default: OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp ENV_OMP_NUM_THREADS=32 notes324_0 = ======================================================================== notes324_1 = 324.apsi_m notes324_2 = OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp notes324_3 = export OMP_NUM_THREADS=32 notes324_4 = 325.apsi_l=peak=default=default: #OPTIMIZE=-O3 -xAVX -ipo1 -openmp #ENV_KMP_AFFINITY=scatter,0 #ENV_OMP_NUM_THREADS=256 basepeak=1 notes325_0 = ======================================================================== notes325_1 = 325.apsi_l notes325_2 = basepeak=1 notes325_6 = 326.gafort_m=peak=default=default: srcalt = ompl.32 ENV_KMP_AFFINITY=scatter,0 notes326_0 = ======================================================================== notes326_1 = 326.gafort_m notes326_2 = srcalt = ompl.32 notes326_3 = export KMP_AFFINITY=scatter,0 notes326_4 = 327.gafort_l=peak=default=default: OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp -rcd -mcmodel=medium -shared-intel ENV_KMP_AFFINITY=scatter,0 ENV_OMP_NUM_THREADS=448 notes327_0 = ======================================================================== notes327_1 = 327.gafort_l notes327_2 = OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp -rcd -mcmodel=medium -shared-intel notes327_4 = export KMP_AFFINITY=scatter,0 notes327_5 = export OMP_NUM_THREADS=448 notes327_6 = 328.fma3d_m=peak=default=default: FOPTIMIZE=-no-prec-sqrt -fp-model fast=2 srcalt = ompl.32 notes328_0 = ======================================================================== notes328_1 = 328.fma3d_m notes328_2 = FOPTIMIZE=-no-prec-sqrt -fp-model fast=2 notes328_3 = srcalt = ompl.32 notes328_4 = 329.fma3d_l=peak=default=default: OPTIMIZE=-O3 -xAVX -ipo1 -openmp -rcd ENV_KMP_AFFINITY=scatter,0 notes329_0 = ======================================================================== notes329_1 = 329.fma3d_l notes329_2 = OPTIMIZE=-O3 -xAVX -ipo1 -openmp -rcd notes329_3 = export KMP_AFFINITY=scatter,0 330.art_m=peak=default=default: COPTIMIZE=-ansi-alias notes330_0 = ======================================================================== notes330_1 = 330.art_m notes330_2 = COPTIMIZE=-ansi-alias notes330_3 = 331.art_l=peak=default=default: #COPTIMIZE=-ansi-alias ENV_OMP_NUM_THREADS=256 notes331_0 = ======================================================================== notes331_1 = 331.art_l notes331_3 = export OMP_NUM_THREADS=256 notes331_4 = 332.ammp_m=peak=default=default: OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp notes332_0 = ======================================================================== notes332_1 = 332.ammp_m notes332_2 = OPTIMIZE=-O3 -xSSE4.2 -ipo1 -openmp notes332_3 = default=default=default=default: hw_vendor=SGI hw_model=SGI UV1000 (Intel Xeon X7542, 2.66GHz) hw_cpu=Intel(R) Xeon(R) Processor X7542 hw_cpu_mhz=2667 hw_fpu=Integrated hw_ncpu=384 cores, 6 chips, 6 cores/chip hw_ncpuorder=2-256 chips hw_parallel= hw_pcache=32 KB I + 32 KB D on chip per core hw_scache=256 KB I+D on chip per core hw_tcache=18 MB I+D on chip hw_ocache=N/A hw_memory=2048 GB (512 x 4 GB Quad-rank DDR3-1066 CL7 DIMMs) hw_disk1=36 TB RAID 5 hw_disk2=80 x 450 GB FC (Seagate Cheetah 15K.6) hw_other= sw_os1=SUSE Linux Enterprise Server 11 SP1, sw_os2=Kernel 2.6.32.12-0.7.1.1381.1.PTF-default sw_compiler=Intel(r) C++ Compiler XE, version 12.0 Beta , Build 20100706 sw_compiler1= Intel(r) Visual Fortran Compiler XE, version 12.0 Beta, Build 20100706 sw_compiler2= GNU C Compiler 4.3.4 sw_file=xfs sw_state=Multi-user, run level 3 company_name=SGI machine_name=SGI UV1000 (Intel Xeon X7542 2.66GHz) license_num=HPG0014 tester_name=SGI test_date=Aug-2010 test_site=SGI hw_avail=Jun-2010 sw_avail=Nov-2010 sw_parallel=OpenMP prepared_by=Huiyu Feng