# Invocation command line: # /local/home/SPECACCEL/bin/runspec -c tesla-pgi142 --reportable -T base,peak -i test,train,ref --device=0 --platform=NVIDIA openacc # output_root was not used for this run ############################################################################ #################################################################### # Tester information #################################################################### license_num = 019 prepared_by = Mathew Colgrove (mathew.colgrove@pgroup.com) tester = NVIDIA Corporation test_sponsor = NVIDIA Corporation ###################################################################### # The header section of the config file. Must appear # before any instances of "default=" # # ext = how the binaries you generated will be identified # tune = specify "base" or "peak" or "all" ext = pgi output_format = asc teeout = yes teerunout = yes tune = base,peak use_submit_for_speed=1 #strict_rundir_verify = 0 #output_root = /local/home//SPECACCEL #################################################################### # HOST Hardware information #################################################################### default=default=default=default: hw_avail = Nov-2013 hw_cpu_name = Intel Core i7-3930K hw_cpu_mhz = 3200 hw_cpu_max_mhz = 3800 hw_fpu = Integrated hw_nchips = 1 hw_ncores = 6 hw_ncoresperchip = 6 hw_nthreadspercore = 2 hw_ncpuorder = 1 chip hw_pcache = 32 KB I + 32 KB D on chip per core hw_scache = 256 KB I+D on chip per core hw_tcache = 12 MB I+D on chip per chip hw_ocache = None hw_vendor = ASUS hw_model = ASUS P9X79 Motherboard sw_parallel_other = No hw_temperature_10201011948889_vendor = Digi hw_temperature_10201011948889_setup = Position 5mm above intake fan hw_temperature_10201011948889_serial = WS34682143 hw_temperature_10201011948889_model = DigiWATCHPORT_H hw_temperature_10201011948889_label = Temperature Meter hw_temperature_10201011948889_connection = USB hw_power_10201011948888_vendor = Xitron Technologies, Inc. hw_power_10201011948888_serial = 28011109005 hw_power_10201011948888_model = 2801 hw_power_10201011948888_met_inst = NIST hw_power_10201011948888_label = Power Analyzer hw_power_10201011948888_connection = RS232 via USB-adapter hw_power_10201011948888_cal_org = Micro Precision Calibration, Inc. hw_power_10201011948888_cal_label = 220081222038459 hw_power_10201011948888_cal_date = 02.20.2014 hw_disk = 99x SCSI, 999 GB, 999 RPM hw_memory000 = 8 GB (2 x 4 GB 2Rx4 PC3-14900R-9, running at 1600 hw_memory001 = MHz) hw_other = None #################################################################### # Accelerator Hardware information #################################################################### hw_accel_model = Tesla K40c hw_accel_vendor = NVIDIA hw_accel_name = NVIDIA Tesla K40c hw_accel_type = GPU hw_accel_connect = PCIe 3.0 16x hw_accel_ecc = Yes hw_accel_desc000 = GPU Boost set to use a graphic clock frequency hw_accel_desc001 = of 810 MHz. See notes below. #################################################################### # Power and temperature meter configuration #################################################################### power = yes power_analyzer = 10.20.101.194:8888 temp_meter = 10.20.101.194:8889 mean_anyway = 1 voltage_range = auto idle_current_range = auto current_range = auto #################################################################### # Software information #################################################################### default=default=default=default: sw_avail = Feb-2014 sw_compiler = PGI Accelerator Server Complete, Release 14.2 sw_accel_driver = NVIDIA UNIX x86_64 Kernel Module 319.60 CC = pgcc CXX = pgc++ FC = pgfortran openacc=default=default=default: sw_other = FFTW 3.3.3 hw_psu = 1200 W hw_psu_info = Thermaltake SMART M1200W opencl=default=default=default: sw_other = CUDA 5.5 SDK openacc=base=default=default: OPTIMIZE = -fast -Mfprelaxed FOPTIMIZE = -acc -ta=tesla:cc35,cuda5.5 COPTIMIZE = -acc -ta=tesla:cc35,cuda5.5 openacc=peak=default=default: OPTIMIZE = -fast -Mfprelaxed FOPTIMIZE = -acc -ta=tesla:cc35,cuda5.5 COPTIMIZE = -acc -ta=tesla:cc35,cuda5.5 opencl=default=default=default: CPORTABILITY = -I/opt/cuda-5.5/include/ -lOpenCL CXXPORTABILITY = -I/opt/cuda-5.5/include/ -lOpenCL opencl=base=default=default: OPTIMIZE = -fast -Mfprelaxed opencl=peak=default=default: OPTIMIZE = -fast -Mfprelaxed 101.tpacf=peak=default=default: OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=1024 103.stencil=peak=default=default: basepeak=yes #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=0 #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_1=0 #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_2=0 104.lbm=peak=default=default: OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=32 OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_1=1 OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_2=1 110.fft=peak=default=default: basepeak = yes 112.spmv=peak=default=default: OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=96 114.mriq=peak=default=default: basepeak = yes #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=256 #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_1_0=256 116.histo=peak=default=default: basepeak = yes #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=512 #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_1_0=512 #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_1_1=1 #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_2_0=1024 #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_3_0=512 117.bfs=peak=default=default: OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=64 OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_1_0=64 118.cutcp=default=default=default: CPORTABILITY += -D__GNUC__ 118.cutcp=peak=default=default: basepeak = yes 120.kmeans=peak=default=default: OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=288 121.lavamd=peak=default=default: basepeak = yes #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=128 122.cfd=peak=default=default: #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=192 #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_1_0=192 #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_2_0=192 OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_3_0=288 #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_4_0=192 123.nw=peak=default=default: basepeak=yes 124.hotspot=peak=default=default: basepeak=yes #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=16 125.lud=peak=default=default: OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=32 126.ge=peak=default=default: OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=512 OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_1_0=1 OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_1_1=512 127.srad=peak=default=default: basepeak=yes #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=256 128.heartwall=peak=default=default: basepeak=yes #OPTIMIZE += -DSPEC_ACCEL_WG_SIZE_0_0=256 140.bplustree=peak=default=default: basepeak = yes 350.md=peak=default=default: FOPTIMIZE = -acc -ta=tesla:cc35,cuda5.5,maxregcount:48 351.palm=peak=default=default: PORTABILITY = -DSPEC_HOST_FFTW3 FOPTIMIZE = -acc=noautopar -ta=tesla:cc35,cuda5.5,fastmath -I$(SPEC)/fftw333/include LIBS = -L$(SPEC)/fftw333/lib -lfftw3 352.ep=peak=default=default: FOPTIMIZE = -acc -ta=tesla:cc35,llvm 359.miniGhost=peak=default=default: FOPTIMIZE = -acc -ta=tesla:cc35,cuda5.5,maxregcount:32 COPTIMIZE = -acc -ta=tesla:cc35,cuda5.5,maxregcount:32 359.miniGhost=default=default=default: EXTRA_LDFLAGS += -Mnomain 363.swim=peak=default=default: FOPTIMIZE = -acc -ta=tesla:cc35,cuda5.5,pin # The following settings were obtained by running 'specperl $[top]/Docs/sysinfo' default: notes_plat_sysinfo_000 = Sysinfo program /local/home/SPECACCEL/Docs/sysinfo notes_plat_sysinfo_005 = $Rev: 6874 $ $Date:: 2013-11-20 #$ 0953404ef7e75a5f9bbb534c6de3f831 notes_plat_sysinfo_010 = running on sbe02 Fri Feb 28 16:50:27 2014 notes_plat_sysinfo_015 = notes_plat_sysinfo_020 = This section contains SUT (System Under Test) info as seen by notes_plat_sysinfo_025 = some common utilities. To remove or add to this section, see: notes_plat_sysinfo_030 = http://www.spec.org/accel/Docs/config.html#sysinfo notes_plat_sysinfo_035 = notes_plat_sysinfo_040 = From /proc/cpuinfo notes_plat_sysinfo_045 = model name : Intel(R) Core(TM) i7-3930K CPU @ 3.20GHz notes_plat_sysinfo_050 = 1 "physical id"s (chips) notes_plat_sysinfo_055 = 12 "processors" notes_plat_sysinfo_060 = cores, siblings (Caution: counting these is hw and system dependent. The notes_plat_sysinfo_065 = following excerpts from /proc/cpuinfo might not be reliable. Use with notes_plat_sysinfo_070 = caution.) notes_plat_sysinfo_075 = cpu cores : 6 notes_plat_sysinfo_080 = siblings : 12 notes_plat_sysinfo_085 = physical 0: cores 0 1 2 3 4 5 notes_plat_sysinfo_090 = cache size : 12288 KB notes_plat_sysinfo_095 = notes_plat_sysinfo_100 = From /proc/meminfo notes_plat_sysinfo_105 = MemTotal: 8130700 kB notes_plat_sysinfo_110 = HugePages_Total: 0 notes_plat_sysinfo_115 = Hugepagesize: 2048 kB notes_plat_sysinfo_120 = notes_plat_sysinfo_125 = /usr/bin/lsb_release -d notes_plat_sysinfo_130 = Red Hat Enterprise Linux Server release 6.4 (Santiago) notes_plat_sysinfo_135 = notes_plat_sysinfo_140 = From /etc/*release* /etc/*version* notes_plat_sysinfo_145 = redhat-release: Red Hat Enterprise Linux Server release 6.4 (Santiago) notes_plat_sysinfo_150 = system-release: Red Hat Enterprise Linux Server release 6.4 (Santiago) notes_plat_sysinfo_155 = system-release-cpe: cpe:/o:redhat:enterprise_linux:6server:ga:server notes_plat_sysinfo_160 = notes_plat_sysinfo_165 = uname -a: notes_plat_sysinfo_170 = Linux sbe02 2.6.32-358.el6.x86_64 #1 SMP Tue Jan 29 11:47:41 EST 2013 x86_64 notes_plat_sysinfo_175 = x86_64 x86_64 GNU/Linux notes_plat_sysinfo_180 = notes_plat_sysinfo_185 = run-level 3 Feb 28 16:45 notes_plat_sysinfo_190 = notes_plat_sysinfo_195 = SPEC is set to: /local/home/SPECACCEL notes_plat_sysinfo_200 = Filesystem Type Size Used Avail Use% Mounted on notes_plat_sysinfo_205 = /dev/mapper/VolGroup-lv_home notes_plat_sysinfo_210 = ext4 860G 53G 764G 7% /local notes_plat_sysinfo_215 = Additional information from dmidecode: notes_plat_sysinfo_220 = notes_plat_sysinfo_225 = Warning: Use caution when you interpret this section. The 'dmidecode' program notes_plat_sysinfo_230 = reads system data which is "intended to allow hardware to be accurately notes_plat_sysinfo_235 = determined", but the intent may not be met, as there are frequent changes to notes_plat_sysinfo_240 = hardware, firmware, and the "DMTF SMBIOS" standard. notes_plat_sysinfo_245 = notes_plat_sysinfo_250 = notes_plat_sysinfo_255 = (End of data from sysinfo program) notes_plat_sysinfo_260 = Information from pgaccelinfo notes_plat_sysinfo_265 = CUDA Driver Version: 5050 notes_plat_sysinfo_270 = NVRM version: NVIDIA UNIX x86_64 Kernel Module 319.60 notes_plat_sysinfo_275 = Wed Sep 25 14:28:26 PDT 2013 notes_plat_sysinfo_280 = Device Number: 0 notes_plat_sysinfo_285 = Device Name: Tesla K40c notes_plat_sysinfo_290 = Device Revision Number: 3.5 notes_plat_sysinfo_295 = Global Memory Size: 12079136768 notes_plat_sysinfo_300 = Number of Multiprocessors: 15 notes_plat_sysinfo_305 = Number of SP Cores: 2880 notes_plat_sysinfo_310 = Number of DP Cores: 960 notes_plat_sysinfo_315 = Concurrent Copy and Execution: Yes notes_plat_sysinfo_320 = Total Constant Memory: 65536 notes_plat_sysinfo_325 = Total Shared Memory per Block: 49152 notes_plat_sysinfo_330 = Registers per Block: 65536 notes_plat_sysinfo_335 = Warp Size: 32 notes_plat_sysinfo_340 = Maximum Threads per Block: 1024 notes_plat_sysinfo_345 = Maximum Block Dimensions: 1024, 1024, 64 notes_plat_sysinfo_350 = Maximum Grid Dimensions: 2147483647 x 65535 x 65535 notes_plat_sysinfo_355 = Maximum Memory Pitch: 2147483647B notes_plat_sysinfo_360 = Texture Alignment: 512B notes_plat_sysinfo_365 = Clock Rate: 810 MHz notes_plat_sysinfo_370 = Max. Clock Rate: 875 MHz notes_plat_sysinfo_375 = Execution Timeout: No notes_plat_sysinfo_380 = Integrated Device: No notes_plat_sysinfo_385 = Can Map Host Memory: Yes notes_plat_sysinfo_390 = Compute Mode: default notes_plat_sysinfo_395 = Concurrent Kernels: Yes notes_plat_sysinfo_400 = ECC Enabled: No notes_plat_sysinfo_405 = Memory Clock Rate: 3004 MHz notes_plat_sysinfo_410 = Memory Bus Width: 384 bits notes_plat_sysinfo_415 = L2 Cache Size: 1572864 bytes notes_plat_sysinfo_420 = Max Threads Per SMP: 2048 notes_plat_sysinfo_425 = Async Engines: 2 notes_plat_sysinfo_430 = Unified Addressing: Yes hw_cpu_name = Intel Core i7-3930K hw_disk = 1000 GB Seagate ST1000DM003 7200 RPM SATA hw_nchips = 1 prepared_by = Mathew Colgrove (mathew.colgrove@pgroup.com) sw_file = ext4 sw_os000 = Red Hat Enterprise Linux Server release 6.4 sw_os001 = (Santiago) sw_os002 = 2.6.32-358.el6.x86_64 sw_state = Run level 3 (multi-user) sw_base_ptrsize = 64-bit sw_peak_ptrsize = 64-bit # 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: flagsurl000 = http://www.spec.org/accel/flags/pgi2014_flags.20150303.xml hw_power_10201011948888_setup000 = connected to the single power supply hw_power_10201011948888_setup001 = that powers the system notes_000 = ECC disabled using the command "nvidia-smi -e 0" notes_005 = GPU Boost mode enabled by setting the device to persistant mode: "nvidia-smi -pm 1" notes_010 = and then setting the memory and graphic clock using: "nvidia-smi -ac ,". notes_015 = For this run, the memory clock was not changed from the default 3004 MHz. notes_020 = The graphic clock was set to the frequency of 810 MHz. notes_025 = nvidai-smi -ac 3004,810 notes_030 = Kit built system using a CoolMaster HAF X case