$ /usr/bin/time src/scripts/nemo.bench mode=3 NEMOBENCH: (2021-03-21) bench3566979 : nbody0=3072 nbody1=10240 nbody3=1000000 size=256 clean=1 bsf=0 tmp=bench3566979 Linux XPS13 5.6.0-1048-oem #52-Ubuntu SMP Thu Feb 18 11:37:00 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux Fri 09 Apr 2021 08:44:36 PM EDT Equal CPU time per code, no output files to check Codes: directcode [mkplummer] gyrfalcON hackcode1 [mkorbit] orbint potcode Tstop: 4.28 5.64 7.89 787400.0 6.66 [5sec CPU for each on a 11th Gen Intel(R) Core(TM) i5-1135G7 up to 4.2GHz] .1.23.45. CPU_USAGE directcode : 5.00 5.00 0.00 0.00 0.00 1840028621 CPU_USAGE gyrfalcON : 4.97 4.96 0.00 0.00 0.00 1840029122 CPU_USAGE hackcode1 : 5.01 5.00 0.00 0.00 0.00 1840029619 CPU_USAGE orbint : 5.04 5.04 0.00 0.00 0.00 1840030121 CPU_USAGE potcode : 5.04 5.02 0.01 0.00 0.00 1840030625 Cleaning bench3566979 (use clean=0 to prevent this) 25.07user 0.02system 0:25.09elapsed 100%CPU (0avgtext+0avgdata 7084maxresident)k 72inputs+4040outputs (3major+6909minor)pagefaults 0swapsIf you run this on older or newer machines, you will most likely see a divergence in speed ratios between the different programs.
$ /usr/bin/time src/scripts/nemo.bench NEMOBENCH: (2020-11-22) bench3555175 : nbody0=3072 nbody1=10240 nbody3=1000000 size=256 clean=1 bsf=0 tmp=bench3555175 Linux XPS13 5.6.0-1042-oem #46-Ubuntu SMP Thu Jan 7 00:51:04 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux Wed 27 Jan 2021 10:47:32 PM EST Times: directcode hackcode1 mkplummer gyrfalcON potcode mkspiral ccdmath ccdpot mkorbit orbint Skipping BSF testing. Use bsf=1 to enable this. CPU_USAGE directcode : 2.35 2.35 0.00 0.00 0.00 1750203216 CPU_USAGE gyrfalcON : 1.75 1.75 0.00 0.00 0.00 1750203579 CPU_USAGE hackcode1 : 1.26 1.26 0.00 0.00 0.00 1750203452 CPU_USAGE mkorbit : 0.00 0.00 0.00 0.00 0.00 1750204776 CPU_USAGE mkplummer : 0.01 0.00 0.00 0.00 0.00 1750203578 CPU_USAGE mkspiral : 4.26 3.09 1.16 0.00 0.00 1750203905 CPU_USAGE orbint : 0.64 0.64 0.00 0.00 0.00 1750204777 CPU_USAGE potcode : 1.49 1.49 0.00 0.00 0.00 1750203755 Processor: 11th Gen Intel(R) Core(TM) i5-1135G7 @ 2.40GHz Cleaning bench927512 (use clean=0 to prevent this) 15.08user 1.32system 0:16.40elapsed 100%CPU (0avgtext+0avgdata 174172maxresident)k
$ /usr/bin/time src/scripts/nemo.bench NEMOBENCH: (2020-05-05) bench129749 : nbody0=3072 nbody1=10240 nbody3=1000000 size=256 clean=1 tmp=bench129749 CPU_USAGE ccdpot : 11.94 11.93 0.00 0.00 0.00 1718371453 CPU_USAGE directcode : 4.02 4.01 0.00 0.00 0.00 1718369452 CPU_USAGE gyrfalcON : 2.49 2.48 0.00 0.00 0.00 1718370096 CPU_USAGE hackcode1 : 2.41 2.39 0.01 0.00 0.00 1718369854 CPU_USAGE mkorbit : 0.00 0.00 0.00 0.00 0.00 1718372647 CPU_USAGE mkplummer : 0.01 0.00 0.00 0.00 0.00 1718370095 CPU_USAGE mkspiral : 8.17 5.90 2.23 0.00 0.00 1718370635 CPU_USAGE orbint : 1.16 1.15 0.00 0.00 0.00 1718372647 CPU_USAGE potcode : 2.90 2.89 0.00 0.00 0.00 1718370345 Cleaning bench129749 (use clean=0 to prevent this) 30.83user 2.53system 0:33.39elapsed 99%CPU (0avgtext+0avgdata 174280maxresident)k 0inputs+4413976outputs (0major+509968minor)pagefaults 0swapsalthough this is on my T480 Lenovo laptop (i7-8550U CPU @ 1.80GHz) I have yet to see any desktop being to beat this.
mkspiral - nbody=10000| gyrfalcON - - step=0 eps=0.05 kmax=6 logfile=/dev/null | uns_density - - all | glnemo2 - all play=t
cp $NEMO/inc/max/maxsizes.h $NEMOLIBelse the installation will automagically take care of this.
ar d $NEMOLIB/libnemo.a toolsnemo.o
#FLIBS = -lcrt1.o -lg2c -lcc_dynamic -lSystem FLIBS = -lg2c -lcc_dynamic -lSystemin the $NEMOLIB/makedefs file.
configure --prefix=$NEMO/host/$NEMOHOSTprobably need a new name for this, or else we could abandon the current scheme and put bin,lib,obj,inc->include, sort of how AIPS++ does it. That's for release 3.1 perhaps. What about other common names, like man and share?
cvs import -d -m "YANC 1.0.5" nemo/usr/dehnen/yanc vendor startdocumentation on YancNemo is already present in YancNemo(1)
Back to: NEMO Home page , the What's Up , or the What's Down page.