# Invocation command line: # /mnt/1/SPEC2006/cpu2006/bin/runspec --rate 192 -c AMD.cfg --iterations 3 --flagsurl /mnt/1/SPEC2006/cpu2006/config/x86-open64-424-flags-rate-revC.20100901.xml -o all fp # output_root was not used for this run ############################################################################ # Invocation command line: # /root/cpu2006-1.1/bin/runspec -c amd1002mc-rate-revC.cfg --rate -T all int -o asc,html --nopreenv --note-preenv # output_root was not used for this run ############################################################################ # AMD SPEC CPU2006 V1.1 Rate Configuration File for 64-bit Linux # # CPU2006 Version 1.1 # Compiler name/version: Open64 4.2.4 # Operating system version: 64-bit Linux # Hardware: AMD Opteron (Magny-Cours) # FP Base Pointer Size: 64-bit only # FP Peak Pointer Size: 32/64-bit # INT Base Pointer Size: 32/64-bit # INT Peak Pointer Size: 32/64-bit # Auto Parallization: Yes # # Important! Please run with your stack size set to 'unlimited'. # Failure to do so may cause 483.xalancbmk to get a stack overflow during # execution. # Using csh: unlimit # Using bash: ulimit -s unlimited # # Your LD_LIBRARY_PATH must be set to the location of the dependent # runtime libraries. (Though this should be taken care of by the # preENV_LD_LIBRARY_PATH found in the AMD1002mc-rate-revX.inc file.) # This line should be commented out during builds. # ##################################################################### ext = amd ignore_errors = no tune = base output_format = all flagsurl000 = http://www.spec.org/cpu2006/flags/x86-open64-424-flags-rate-revC.20100901.xml # $[top]/amd-platform-rate-revC.xml size = test,train,ref check_md5 = yes reportable = yes env_vars = no teeout = yes mean_anyway = yes teerunout = yes post_setup = sync ##################################################################### # Machine Specific Settings ##################################################################### # If your machine orders nodes sequentially, you have 32GB/socket # memory using 4GB DIMMS, and your system has huge pages configured, # most settings will be correctly configured based on the below two # settings. See README.amd1002mc-rate for more information. # How many sockets does your test system have? (2 or 4) #%define num_sockets 4 # How many cores per socket does your system have? (8 or 12) #%define cores_per_socket 8 ######################################################################## # You should not need to change anything in this section unless you have # a non-standard configuration. See README.amd1002mc-rate for more # information. ######################################################################## #%if ('%{cores_per_socket}' eq '12') && ('%{num_sockets}' eq '2') #% define num_system_huge_pages 10800 #% define cores_per_node 6 #% define mem_size 64 #% define mem_sticks 16 #%elif ('%{cores_per_socket}' eq '12') && ('%{num_sockets}' eq '4') #% define num_system_huge_pages 21600 #% define cores_per_node 6 #% define mem_size 128 #% define mem_sticks 32 #%elif ('%{cores_per_socket}' eq '8') && ('%{num_sockets}' eq '2') #% define num_system_huge_pages 14336 #% define cores_per_node 4 #% define mem_size 64 #% define mem_sticks 16 #%elif ('%{cores_per_socket}' eq '8') && ('%{num_sockets}' eq '4') #% define num_system_huge_pages 28672 #% define cores_per_node 4 #% define mem_size 128 #% define mem_sticks 32 #%endif #################################################################### # Tester information #################################################################### license_num = 2929 prepared_by = ScaleMP tester = ScaleMP test_sponsor = ScaleMP hw_vendor = ScaleMP hw_model000 = vSMP Foundation (AMD Opteron 6172, 2.10 GHz) #################################################################### # Hardware information #################################################################### hw_avail = Aug-2010 hw_cpu_name = AMD Opteron 6172 hw_cpu_mhz = 2100 hw_fpu = Integrated hw_nchips = 16 hw_ncores = 192 hw_ncoresperchip = 12 hw_nthreadspercore = 1 hw_ncpuorder = 2 to 64 boards with 2 to 4 chips per board hw_pcache = 64 KB I + 64 KB D on chip per core hw_scache = 512 KB I+D on chip per core hw_tcache = 12 MB I+D on chip per chip, 6 MB shared / 6 cores hw_ocache = 60 GB I+D off chip per system hw_disk = 8 x 250 GB SATA, 7200 RPM hw_memory = 512 GB (8 x 16 x 4 GB 2Rx4 PC3-10600R-9, ECC) hw_other = None sw_file = xfs sw_os000 = Red Hat Enterprise Linux Server release 5.5 sw_os001 = (Tikanga) sw_os002 = Kernel 2.6.32.46-5.vSMP sw_state = Run level 3 (multi-user) ##################################################################### # Notes ##################################################################### notes_000 =The x86 Open64 Compiler Suite is only available from (and supported by) AMD at notes_005 =http://developer.amd.com/cpu/open64 notes_010 =Binaries were compiled on RHEL5.5 with binutils 2.18 notes_os_000 ='ulimit -s unlimited' was used to set environment stack size notes_os_005 ='ulimit -l 2097152' was used to set environment locked pages in memory limit # submit = echo "$command" > run.sh ; $BIND bash run.sh # submit = taskset -c $SPECCOPYNUM $command submit = taskset -c \$((`grep -wc processor /proc/cpuinfo` - $SPECCOPYNUM - 1)) $command ############################ Software Info ############################ fp=default=default=default: sw_base_ptrsize = 64-bit sw_peak_ptrsize = Not Applicable sw_avail = Sep-2011 sw_compiler000 = C/C++/Fortran: Version 4.2.4 of Open64 Compiler sw_compiler001 = Suite int=default=default=default: sw_base_ptrsize = 32/64-bit sw_peak_ptrsize = 32/64-bit sw_avail = Jul-2010 sw_compiler = x86 Open64 4.2.4 Compiler Suite (from AMD) sw_other001 = Hoard (libhoard) 3.8, 32-bit Library for Linux default=default=default=default: ##################################################################### # Compiler selection # # CC = opencc -march=barcelona -mso CXX = openCC -march=barcelona -mso FC = openf95 -march=barcelona -mso ##################################################################### # Portability for default 64-bit code generation ##################################################################### default=default=default: PORTABILITY = -DSPEC_CPU_LP64 ##################################################################### # INT Portability ##################################################################### 400.perlbench=default=default: CPORTABILITY = -DSPEC_CPU_LINUX_X64 403.gcc=peak=default: PORTABILITY = 429.mcf=peak=default: PORTABILITY = 462.libquantum=default=default: CPORTABILITY = -DSPEC_CPU_LINUX 471.omnetpp=default=default: # Needed to avoid -DSPEC_CPU_LP64 on -m32 C++ base codes PORTABILITY = 473.astar=default=default: # Needed to avoid -DSPEC_CPU_LP64 on -m32 C++ base codes PORTABILITY = 483.xalancbmk=default=default: CXXPORTABILITY = -DSPEC_CPU_LINUX # Needed to avoid -DSPEC_CPU_LP64 on -m32 C++ base codes PORTABILITY = ################################################################# # # # SPECint Tuning & Notes # # # ################################################################# int=base: #COPTIMIZE = -Ofast -CG:local_sched_alg=1 -INLINE:aggressive=on -IPA:plimit=8000:small_pu=100 -HP:bdt=2m:heap=2m COPTIMIZE = -Ofast -CG:local_sched_alg=1 -INLINE:aggressive=on -IPA:plimit=8000:small_pu=100 CXXOPTIMIZE = -Ofast -m32 -INLINE:aggressive=on -CG:cmp_peep=on EXTRA_CXXLIBS = /usr/local/lib/libhoard.so ##################################################################### # INT Peak Tuning ##################################################################### int=peak=default: #COPTIMIZE = -Ofast -CG:local_sched_alg=1 -HP:bdt=2m:heap=2m COPTIMIZE = -Ofast -CG:local_sched_alg=1 #CXXOPTIMIZE = -Ofast -m32 -HP:bdt=2m:heap=2m CXXOPTIMIZE = -Ofast -m32 PASS1_CFLAGS = -fb_create fbdata PASS1_CXXFLAGS = -fb_create fbdata PASS1_LDFLAGS = -fb_create fbdata PASS2_CFLAGS = -fb_opt fbdata PASS2_CXXFLAGS = -fb_opt fbdata PASS2_LDFLAGS = -fb_opt fbdata 400.perlbench=peak=default: #COPTIMIZE = -Ofast -IPA:plimit=20000 -LNO:opt=0 -OPT:unroll_times_max=8:unroll_size=256:unroll_level=2:keep_ext=on -WOPT:if_conv=0 -CG:local_sched_alg=1:unroll_fb_req=on -HP:bdt=2m:heap=2m COPTIMIZE = -Ofast -IPA:plimit=20000 -LNO:opt=0 -OPT:unroll_times_max=8:unroll_size=256:unroll_level=2:keep_ext=on -WOPT:if_conv=0 -CG:local_sched_alg=1:unroll_fb_req=on 401.bzip2=peak=default: #COPTIMIZE = -O3 -OPT:alias=disjoint:goto=off -CG:local_sched_alg=1 -HP:bdt=2m:heap=2m COPTIMIZE = -O3 -OPT:alias=disjoint:goto=off -CG:local_sched_alg=1 403.gcc=peak=default: #COPTIMIZE = -Ofast -LNO:trip_count=256:prefetch_ahead=10 -CG:cmp_peep=on -m32 -HP:bdt=2m:heap=2m -GRA:unspill=on -IPA:small_pu=200 COPTIMIZE = -Ofast -LNO:trip_count=256:prefetch_ahead=10 -CG:cmp_peep=on -m32 -GRA:unspill=on -IPA:small_pu=200 429.mcf=peak=default: #COPTIMIZE = -O3 -ipa -INLINE:aggressive=on -CG:gcm=off -GRA:prioritize_by_density=on -m32 -HP:bdt=2m:heap=2m COPTIMIZE = -O3 -ipa -INLINE:aggressive=on -CG:gcm=off -GRA:prioritize_by_density=on -m32 feedback = 0 445.gobmk=peak=default: #COPTIMIZE = -O3 -OPT:alias=restrict:unroll_times_max=8:unroll_size=256:unroll_level=2:keep_ext=on -ipa -IPA:plimit=750:min_hotness=300:pu_reorder=1 -LNO:prefetch=1:ignore_feedback=off -CG:p2align=on:unroll_fb_req=on -HP:bdt=2m:heap=2m COPTIMIZE = -O3 -OPT:alias=restrict:unroll_times_max=8:unroll_size=256:unroll_level=2:keep_ext=on -ipa -IPA:plimit=750:min_hotness=300:pu_reorder=1 -LNO:prefetch=1:ignore_feedback=off -CG:p2align=on:unroll_fb_req=on 456.hmmer=peak=default: #COPTIMIZE = -Ofast -LNO:prefetch=0 -OPT:alias=disjoint:unroll_times_max=8:unroll_size=256:unroll_level=2:keep_ext=on -CG:local_sched_alg=1:cflow=0:push_pop_int_saved_regs=off:cmp_peep=on -HP:bdt=2m:heap=2m COPTIMIZE = -Ofast -LNO:prefetch=0 -OPT:alias=disjoint:unroll_times_max=8:unroll_size=256:unroll_level=2:keep_ext=on -CG:local_sched_alg=1:cflow=0:push_pop_int_saved_regs=off:cmp_peep=on 458.sjeng=peak=default: COPTIMIZE = -O3 -ipa -LNO:ignore_feedback=off:full_unroll=10:fusion=0:fission=2 -IPA:pu_reorder=2 -CG:ptr_load_use=0 -OPT:unroll_times_max=8 -INLINE:aggressive=on 462.libquantum=peak=default: #COPTIMIZE = -Ofast -LNO:pf2=0 -CG:gcm=off:use_prefetchnta=on:cmp_peep=on -WOPT:aggstr=0 -HP:bdt=2m:heap=2m -OPT:alias=disjoint -INLINE:aggressive=on -IPA:space=1000:plimit=20000 COPTIMIZE = -Ofast -LNO:pf2=0 -CG:gcm=off:use_prefetchnta=on:cmp_peep=on -WOPT:aggstr=0 -OPT:alias=disjoint -INLINE:aggressive=on -IPA:space=1000:plimit=20000 feedback = 0 464.h264ref=peak=default: COPTIMIZE = -O3 -IPA:plimit=20000 -OPT:alias=disjoint -LNO:prefetch=0 -CG:ptr_load_use=0:push_pop_int_saved_regs=off 471.omnetpp=peak=default: basepeak = 1 473.astar=peak=default: #CXXOPTIMIZE = -Ofast -TENV:frame_pointer=off -WOPT:if_conv=0 -GRA:optimize_boundary=on -OPT:alias=disjoint -INLINE:aggressive=on -IPA:small_pu=3000:plimit=3000 -m32 -HP:bdt=2m:heap=2m CXXOPTIMIZE = -Ofast -TENV:frame_pointer=off -WOPT:if_conv=0 -GRA:optimize_boundary=on -OPT:alias=disjoint -INLINE:aggressive=on -IPA:small_pu=3000:plimit=3000 -m32 483.xalancbmk=peak=default: CXXOPTIMIZE = -Ofast -INLINE:aggressive=on -m32 -CG:cmp_peep=on -GRA:unspill=on -TENV:frame_pointer=off -fno-emit-exceptions EXTRA_CXXLIBS = -L$(SMARTHEAP_DIR) -lsmartheap feedback = 0 ##################################################################### # FP Portability ##################################################################### 436.cactusADM=default=default: FPORTABILITY = -fno-second-underscore 447.dealII=peak=default: PORTABILITY = 450.soplex=peak=default: PORTABILITY = 481.wrf=default=default: CPORTABILITY = -DSPEC_CPU_LINUX -DSPEC_CPU_CASE_FLAG FPORTABILITY = -fno-second-underscore ################################################################# # # # SPECfp Tuning # # # ################################################################# fp=base: #COPTIMIZE = -Ofast -OPT:malloc_alg=1 -HP:bdt=2m COPTIMIZE = -Ofast -OPT:malloc_alg=1 #CXXOPTIMIZE = -Ofast -static -INLINE:aggressive=on -OPT:malloc_alg=1 -HP:bdt=2m CXXOPTIMIZE = -Ofast -static -INLINE:aggressive=on -OPT:malloc_alg=1 #FOPTIMIZE = -Ofast -HP FOPTIMIZE = -Ofast ##################################################################### # FP Peak Tuning ##################################################################### fp=peak=default: #COPTIMIZE = -Ofast -HP:bdt=2m:heap=2m COPTIMIZE = -Ofast #CXXOPTIMIZE = -Ofast -INLINE:aggressive=on -HP:bdt=2m:heap=2m CXXOPTIMIZE = -Ofast -INLINE:aggressive=on #FOPTIMIZE = -Ofast -HP:bdt=2m:heap=2m FOPTIMIZE = -Ofast PASS1_CFLAGS = -fb_create fbdata PASS2_CFLAGS = -fb_opt fbdata PASS1_CXXFLAGS = -fb_create fbdata PASS2_CXXFLAGS = -fb_opt fbdata PASS1_FFLAGS = -fb_create fbdata PASS2_FFLAGS = -fb_opt fbdata PASS1_LDFLAGS = -fb_create fbdata PASS2_LDFLAGS = -fb_opt fbdata 410.bwaves=peak=default: #FOPTIMIZE = -O3 -OPT:Ofast:treeheight=on -LNO:blocking=off:prefetch_ahead=5:ignore_feedback=off -WOPT:aggstr=0 -HP:bdt=2m:heap=2m -CG:cmp_peep=on FOPTIMIZE = -O3 -OPT:Ofast:treeheight=on -LNO:blocking=off:prefetch_ahead=5:ignore_feedback=off -WOPT:aggstr=0 -CG:cmp_peep=on feedback = 0 416.gamess=peak=default: #FOPTIMIZE = -O3 -LNO:fu=6:blocking=0:prefetch=0 -OPT:Ofast:ro=3:unroll_size=256 -HP:bdt=2m:heap=2m FOPTIMIZE = -O3 -LNO:fu=6:blocking=0:prefetch=0 -OPT:Ofast:ro=3:unroll_size=256 433.milc=peak=default: #COPTIMIZE = -Ofast -CG:movnti=1:local_sched_alg=1 -CG:locs_shallow_depth=1 -HP:bdt=2m:heap=2m -LNO:prefetch=3 COPTIMIZE = -Ofast -CG:movnti=1:local_sched_alg=1 -CG:locs_shallow_depth=1 -LNO:prefetch=3 feedback = 0 434.zeusmp=peak=default: #FOPTIMIZE = -Ofast -LNO:blocking=off:interchange=off -OPT:treeheight=on:unroll_size=256 -CG:cmp_peep=on -GRA:prioritize_by_density=on -HP FOPTIMIZE = -Ofast -LNO:blocking=off:interchange=off -OPT:treeheight=on:unroll_size=256 -CG:cmp_peep=on -GRA:prioritize_by_density=on feedback = 0 435.gromacs=peak=default: #FOPTIMIZE = -Ofast -OPT:rsqrt=2 -HP:bdt=2m:heap=2m FOPTIMIZE = -Ofast -OPT:rsqrt=2 COPTIMIZE = $(FOPTIMIZE) feedback = 0 436.cactusADM=peak=default: #FOPTIMIZE = -Ofast -apo -LNO:prefetch_ahead=1 -HP:bdt=2m:heap=2m -LANG:heap_allocation_threshold=100 FOPTIMIZE = -Ofast -apo -LNO:prefetch_ahead=1 -LANG:heap_allocation_threshold=100 COPTIMIZE = $(FOPTIMIZE) 437.leslie3d=peak=default: #FOPTIMIZE = -Ofast -HP:bdt=2m:heap=2m FOPTIMIZE = -Ofast feedback = 0 444.namd=peak=default: #CXXOPTIMIZE = -Ofast -LNO:ignore_feedback=off -CG:local_sched_alg=2:load_exe=0:compute_to=on -OPT:unroll_size=256 -fno-exceptions -HP:bdt=2m:heap=2m CXXOPTIMIZE = -Ofast -LNO:ignore_feedback=off -CG:local_sched_alg=2:load_exe=0:compute_to=on -OPT:unroll_size=256 -fno-exceptions 447.dealII=peak=default: #CXXOPTIMIZE = -Ofast -static -INLINE:aggressive=on -LNO:opt=0 -fno-emit-exceptions -m32 -OPT:unroll_times_max=8:unroll_size=256:unroll_level=2 -HP:bdt=2m:heap=2m -GRA:unspill=on -CG:cmp_peep=on -TENV:frame_pointer=off CXXOPTIMIZE = -Ofast -static -INLINE:aggressive=on -LNO:opt=0 -fno-emit-exceptions -m32 -OPT:unroll_times_max=8:unroll_size=256:unroll_level=2 -GRA:unspill=on -CG:cmp_peep=on -TENV:frame_pointer=off feedback = 0 450.soplex=peak=default: #CXXOPTIMIZE = -O3 -INLINE:aggressive=on -OPT:IEEE_arith=3:IEEE_NaN_Inf=off:fold_unsigned_relops=on:malloc_alg=1 -CG:load_exe=0 -fno-exceptions -m32 -HP:bdt=2m CXXOPTIMIZE = -O3 -INLINE:aggressive=on -OPT:IEEE_arith=3:IEEE_NaN_Inf=off:fold_unsigned_relops=on:malloc_alg=1 -CG:load_exe=0 -fno-exceptions -m32 453.povray=peak=default: CXXOPTIMIZE = -Ofast -INLINE:aggressive=on 454.calculix=peak=default: #FOPTIMIZE = -Ofast -CG:load_exe=0:ptr_load_use=0:local_sched_alg=2:compute_to=on -LNO:prefetch_ahead=30 -WOPT:unroll=2 -GRA:optimize_boundary=on -HP:bdt=2m:heap=2m FOPTIMIZE = -Ofast -CG:load_exe=0:ptr_load_use=0:local_sched_alg=2:compute_to=on -LNO:prefetch_ahead=30 -WOPT:unroll=2 -GRA:optimize_boundary=on COPTIMIZE = $(FOPTIMIZE) feedback = 0 459.GemsFDTD=peak=default: #FOPTIMIZE = -Ofast -LNO:fission=2:prefetch_ahead=1 -CG:load_exe=0:local_sched_alg=1 -HP FOPTIMIZE = -Ofast -LNO:fission=2:prefetch_ahead=1 -CG:load_exe=0:local_sched_alg=1 feedback = 0 465.tonto=peak=default: #FOPTIMIZE = -Ofast -OPT:alias=no_f90_pointer_alias -LNO:blocking=off -CG:load_exe=1 -IPA:plimit=525 -HP FOPTIMIZE = -Ofast -OPT:alias=no_f90_pointer_alias -LNO:blocking=off -CG:load_exe=1 -IPA:plimit=525 feedback = 0 470.lbm=peak=default: basepeak = 1 481.wrf=peak=default: #FOPTIMIZE = -Ofast -LNO:blocking=off:prefetch_ahead=10 -LANG:copyinout=off -IPA:callee_limit=5000 -GRA:prioritize_by_density=on -m3dnow -HP FOPTIMIZE = -Ofast -LNO:blocking=off:prefetch_ahead=10 -LANG:copyinout=off -IPA:callee_limit=5000 -GRA:prioritize_by_density=on -m3dnow COPTIMIZE = $(FOPTIMIZE) feedback = 0 482.sphinx3=peak=default: COPTIMIZE = -Ofast -OPT:malloc_alg=2 -CG:sse_cse_regs=0:locs_shallow_depth=1:cmp_peep=on:local_sched_alg=1 -INLINE:aggressive=on # 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. 462.libquantum: # The following setting was inserted automatically as a result of # post-run basepeak application. basepeak = 1 # 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_plat_000 = ScaleMP notes_plat_005 = vSMP Foundation: 3.5.284, released Sep. 13, 2011 notes_plat_010 = See http://www.scalemp.com notes_plat_015 = Hardware Details: notes_plat_020 = System was aggregated using 8 X SuperMicro AS-2122TG-HIBQRF servers. notes_plat_025 = The servers were connected with Melanox InfiniBand QDR and a notes_plat_030 = QDR switch. sw_other001 = ScaleMP vSMP Foundation 3.5.284, released Sep. sw_other002 = 13, 2011 notes_submit_000 =The taskset command was used to bind processes to cores.