# Invocation command line: # /home/pshelepu/mpi-2007-l0224/bin/runspec --reportable --config intel_mpi2007-224.cfg --flagsurl EM64T_Intel140_flags.xml --size lref --iterations 3 --output_format=txt --ranks=224 large # output_root was not used for this run ############################################################################ ##################################################################### # # Config file to run SPEC MPI2007 with Intel Software Toolchain # (Intel Compiler 14.0.3 and Intel MPI 4.1.3) # ##################################################################### env_vars = 1 basepeak = 1 reportable = 1 ignore_errors = 0 iterations = 3 makeflags = -j 4 tune = base size = mref use_version_url = 1 version_url = version.txt use_submit_for_speed = 1 output_format = all FC = mpiifort CC = mpiicc CXX = mpiicpc ##################################################################### # Portability flags ##################################################################### 121.pop2=default=default=default: CPORTABILITY = -DSPEC_MPI_CASE_FLAG 126.lammps=default=default=default: CXXPORTABILITY = -DMPICH_IGNORE_CXX_SEEK 127.wrf2=default=default=default: CPORTABILITY = -DSPEC_MPI_CASE_FLAG -DSPEC_MPI_LINUX 130.socorro=default=default=default: srcalt=nullify_ptrs FPORTABILITY = -assume nostd_intent_in ################################################################# # Optimization flags ################################################################# default=default=default=default: OPTIMIZE = -O3 -xCORE-AVX2 -no-prec-div submit = mpiexec.hydra -machinefile \$LSB_DJOB_HOSTFILE -genv I_MPI_DEVICE rdssm -genv I_MPI_FALLBACK_DEVICE disable -n $ranks $command ################################################################# # Notes ################################################################# test_sponsor = Intel Corporation license_num = 13 tester = Pavel Shelepugin test_date = Aug-2013 hw_avail = Sep-2014 sw_avail = May-2014 prepared_by = Intel Corporation system_vendor = Intel Corporation system_name000 = Endeavor (Intel Xeon E5-2697 v3, 2.60 GHz, system_name001 = DDR4-2133 MHz, SMT on, Turbo off) # # Computation node info # node_compute_label = Endeavor Node node_compute_order = 1 node_compute_count = 8 node_compute_purpose = compute node_compute_hw_vendor = Intel node_compute_hw_model = R2208WTTYC1 node_compute_hw_cpu_name = Intel Xeon E5-2697 v3 node_compute_hw_ncpuorder = 1-2 chips node_compute_hw_nchips = 2 node_compute_hw_ncores = 28 node_compute_hw_ncoresperchip = 14 node_compute_hw_nthreadspercore = 2 node_compute_hw_cpu_char000 = Intel Turbo Boost Technology disabled, node_compute_hw_cpu_char001 = 9.6 GT/s QPI, Hyper-Threading enabled node_compute_hw_cpu_mhz = 2600 node_compute_hw_pcache = 32 KB I + 32 KB D on chip per core node_compute_hw_scache = 256 KB I+D on chip per core node_compute_hw_tcache000= 35 MB I+D on chip per chip, 35 MB node_compute_hw_tcache001 = shared / 14 cores node_compute_hw_ocache = None node_compute_hw_memory = 64 GB (8 x 8 GB 2Rx4 PC4-17000R-15, ECC) node_compute_hw_disk = ATA INTEL SSDSA2BZ20, SSDSC2BB80 node_compute_hw_other = None node_compute_hw_adapter_ib_model = Mellanox MCX353A-FCAT ConnectX-3 node_compute_hw_adapter_ib_count = 1 node_compute_hw_adapter_ib_slot_type = PCIe x8 Gen3 node_compute_hw_adapter_ib_data_rate = InfiniBand 4x FDR node_compute_hw_adapter_ib_ports_used = 1 node_compute_hw_adapter_ib_interconnect = InfiniBand node_compute_hw_adapter_ib_driver = OFED 3.5-2-MIC-rc1 node_compute_hw_adapter_ib_firmware = 2.31.5050 node_compute_hw_adapter_fs_model000= Intel (ESB2) 82575EB Dual-Port Gigabit node_compute_hw_adapter_fs_model001 = Ethernet Controller node_compute_hw_adapter_fs_count = 1 node_compute_hw_adapter_fs_slot_type = PCI-Express x8 node_compute_hw_adapter_fs_data_rate = 1Gbps Ethernet node_compute_hw_adapter_fs_ports_used = 2 node_compute_hw_adapter_fs_interconnect = Ethernet node_compute_hw_adapter_fs_driver = e1000 node_compute_hw_adapter_fs_firmware = None node_compute_sw_os = Red Hat EL 6.5, kernel 2.6.32-358 node_compute_sw_localfile = Linux/xfs node_compute_sw_sharedfile = NFS node_compute_sw_state = Multi-User node_compute_sw_other = IBM Platform LSF Standard 9.1.1.1 # # Fileserver node info # node_fileserver_label = NFS node_fileserver_order = 2 node_fileserver_count = 1 node_fileserver_purpose = fileserver node_fileserver_hw_vendor = Intel node_fileserver_hw_model = S7000FC4UR node_fileserver_hw_cpu_name = Intel Xeon CPU node_fileserver_hw_ncpuorder = 1-4 chips node_fileserver_hw_nchips = 4 node_fileserver_hw_ncores = 16 node_fileserver_hw_ncoresperchip = 4 node_fileserver_hw_nthreadspercore = 2 node_fileserver_hw_cpu_char = -- node_fileserver_hw_cpu_mhz = 2926 node_fileserver_hw_pcache = 32 KB I + 32 KB D on chip per core node_fileserver_hw_scache = 8 MB I+D on chip per chip, 4 MB shared / 2 cores node_fileserver_hw_tcache = None node_fileserver_hw_ocache = None node_fileserver_hw_memory = 64 GB node_fileserver_hw_disk = 8 disks, 500GB/disk, 2.7TB total node_fileserver_hw_other = None node_fileserver_hw_adapter_fs_model000 = Intel 82563GB Dual-Port Gigabit node_fileserver_hw_adapter_fs_model001 = Ethernet Controller node_fileserver_hw_adapter_fs_count = 1 node_fileserver_hw_adapter_fs_slot_type = PCI-Express x8 node_fileserver_hw_adapter_fs_data_rate = 1Gbps Ethernet node_fileserver_hw_adapter_fs_ports_used = 1 node_fileserver_hw_adapter_fs_interconnect = Ethernet node_fileserver_hw_adapter_fs_driver = e1000e node_fileserver_hw_adapter_fs_firmware = N/A node_fileserver_sw_os = RedHat EL 5 Update 4 node_fileserver_sw_localfile = None node_fileserver_sw_sharedfile = NFS node_fileserver_sw_state = Multi-User node_fileserver_sw_other = None # # IB interconnect # interconnect_ib_label = IB Switch interconnect_ib_order = 1 interconnect_ib_purpose = MPI traffic interconnect_ib_hw_vendor = Mellanox interconnect_ib_hw_model = Mellanox MSX6025F-1BFR interconnect_ib_hw_switch_3600_model = Mellanox MSX6025F-1BFR interconnect_ib_hw_switch_3600_count = 46 interconnect_ib_hw_switch_3600_ports = 36 interconnect_ib_hw_topo = Fat tree interconnect_ib_hw_switch_3600_data_rate = InfiniBand 4x FDR interconnect_ib_hw_switch_3600_firmware = 9.2.8000 # # Cluster file system interconnect # interconnect_fs_label = Gigabit Ethernet interconnect_fs_order = 2 interconnect_fs_purpose = Cluster File System interconnect_fs_hw_vendor = Force10 Networks, Cisco Systems interconnect_fs_hw_model = Force10 S50N, Force10 C300, Cisco WS-C4948E-F interconnect_fs_hw_switch_fs_model = Force10 S50N, Force10 C300, Cisco WS-C4948E-F interconnect_fs_hw_switch_fs_count = 13 interconnect_fs_hw_switch_fs_ports = 48 interconnect_fs_hw_topo = Star interconnect_fs_hw_switch_fs_data_rate = 1Gbps Ethernet, 10Gbps Ethernet interconnect_fs_hw_switch_fs_firmware = 8.3.2.0, 12.2(54)WO # # Hardware # system_class = Homogeneous max_ranks = 224 max_peak_ranks = 224 # # Software # sw_c_compiler000= Intel C++ Composer XE 2013 for Linux, Version sw_c_compiler001 = 14.0.3.174 Build 20140422 sw_cxx_compiler000= Intel C++ Composer XE 2013 for Linux, Version sw_cxx_compiler001 = 14.0.3.174 Build 20140422 sw_f_compiler000= Intel Fortran Composer XE 2013 for Linux, Version sw_f_compiler001 = 14.0.3.174 Build 20140422 sw_auto_parallel = sw_base_ptrsize = 64-bit sw_peak_ptrsize = 64-bit sw_mpi_library = Intel MPI Library 4.1.3.049 for Linux sw_mpi_other = None sw_preprocessors = No sw_other = None # # General notes # notes_000 = MPI startup command: notes_005 = mpiexec.hydra command was used to start MPI jobs. notes_010 = notes_015 = BIOS settings: notes_020 = Intel Hyper-Threading Technology (SMT): Enabled (default is Enabled) notes_025 = Intel Turbo Boost Technology (Turbo) : Disabled (default is Enabled) notes_030 = notes_035 = RAM configuration: notes_040 = Compute nodes have 2x8-GB RDIMM on each memory channel. notes_045 = notes_050 = Network: notes_055 = Forty six 36-port switches: 18 core switches and 28 leaf switches. notes_060 = Each leaf has one link to each core. Remaining 18 ports on 25 of 28 leafs notes_065 = are used for compute nodes. On the remaining 3 leafs the ports are used notes_070 = for FS nodes and other peripherals. notes_075 = notes_080 = Job placement: notes_085 = Each MPI job was assigned to a topologically compact set of nodes, i.e. notes_090 = the minimal needed number of leaf switches was used for each job: 1 switch notes_095 = for 28/56/112/224/448 ranks, 2 switches for 896 ranks, 4 switches for 1792 ranks, notes_100 = 8 switches for 3584 ranks. notes_105 = notes_110 = IBM Platform LSF was used for job submission. It has no impact on performance. notes_115 = Information can be found at: http://www.ibm.com # 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/mpi2007/flags/EM64T_Intel140_flags.20140908.xml