n-body benchmark N=50,000,000

Each chart bar shows how many times more Memory, one ↓ n-body program used, compared to the program that used least Memory.

These are not the only programs that could be written. These are not the only compilers and interpreters. These are not the only programming languages.

Column × shows how many times more each program used compared to the benchmark program that used least.

    sortsort sort
  ×   Program Source Code CPU secs Elapsed secs Memory KB Code B ≈ CPU Load
1.0Pascal Free Pascal 34.7534.7681308  1% 0% 1% 100%
1.0Pascal Free Pascal #2 37.0137.0281322  1% 0% 0% 100%
1.0Pascal Free Pascal #3 35.4035.4181418  1% 1% 0% 100%
45C++ g++ #4 20.9320.943601428  1% 1% 0% 100%
45C gcc #5 20.9320.943601429  1% 0% 0% 100%
54C gcc #4 10.1910.194321490  1% 0% 0% 100%
54C++ g++ #6 19.3419.354321668  1% 1% 1% 100%
54C gcc #6 20.5920.604321180  1% 0% 1% 100%
55C gcc #2 22.2022.214361263  0% 1% 1% 100%
55C++ g++ 19.5419.544361659  1% 0% 1% 100%
55C gcc 20.8120.824361173  0% 1% 1% 100%
55C++ g++ #8 9.359.354361544  1% 1% 1% 100%
55C++ g++ #7 9.799.794361545  0% 1% 1% 100%
55C gcc #3 20.6620.674361208  1% 0% 0% 100%
64Fortran Intel 20.0620.065121389  0% 0% 0% 100%
64Fortran Intel #3 21.5321.545121299  0% 0% 0% 100%
64Fortran Intel #4 22.4722.485121172  0% 0% 0% 100%
64Fortran Intel #5 9.059.055121659  1% 0% 1% 100%
64Fortran Intel #2 19.9519.965121496  0% 0% 0% 100%
90OCaml 21.8621.877201239  0% 1% 0% 100%
104C++ g++ #3 9.929.928281763  0% 1% 0% 100%
130Go 22.9622.961,0361310  1% 0% 0% 100%
130Lua 8 min8 min1,0361201  0% 0% 0% 100%
130Lua #2 7 min7 min1,0401193  0% 0% 0% 100%
130Lua #4 7 min7 min1,0401305  0% 0% 0% 100%
134C++ g++ #5 13.4513.451,0681749  1% 0% 1% 100%
187Ada 2005 GNAT #5 15.4315.441,4922427  1% 0% 0% 100%
188Ada 2005 GNAT #3 23.2723.281,5001740  1% 0% 0% 100%
188Ada 2005 GNAT 25.9225.931,5001608  1% 0% 0% 100%
293Haskell GHC #2 22.6522.662,3441874  0% 1% 1% 100%
295Haskell GHC 27.4027.412,3601706  1% 0% 0% 100%
328Perl #2 17 min17 min2,6201401  0% 0% 0% 100%
418PHP #3 11 min11 min3,3401082  0% 0% 0% 100%
598Rust 21.3321.344,7841371  0% 1% 0% 100%
697Python 3 15 min15 min5,5721181  1% 0% 0% 100%
940Ruby #2 10 min10 min7,5201137  0% 0% 0% 100%
1,082Lisp SBCL 42.8542.878,6561363  1% 0% 0% 100%
1,082Lisp SBCL #2 26.6726.688,6561398  0% 1% 0% 100%
1,221JavaScript V8 #2 37.1937.219,7641527  1% 0% 0% 100%
1,279JavaScript V8 37.9938.0110,2321287  1% 0% 0% 100%
1,514Erlang HiPE #3 119.80119.8412,1081399  0% 0% 0% 100%
2,241Dart #2 30.7030.7117,9281274  1% 1% 0% 100%
2,347Dart #3 28.3528.3618,7761420  0% 1% 0% 100%
2,446Java  #2 22.6622.6719,5681424  0% 0% 0% 100%
2,749C# Mono #6 24.4624.4721,9921289  0% 0% 0% 100%
2,763C# Mono #2 22.6722.6822,1041410  1% 1% 1% 100%
2,788C# Mono #3 21.7521.7622,3041305  1% 1% 0% 100%
2,794C# Mono #4 30.8130.8222,3521461  1% 0% 1% 100%
2,801C# Mono 37.9737.9822,4041403  1% 0% 0% 100%
2,822C# Mono #7 28.1428.1522,5721539  1% 0% 1% 100%
2,846C# Mono #5 31.5931.6022,7641485  1% 0% 0% 100%
3,255Scala 26.4826.4926,0401358  0% 0% 0% 100%
3,441Racket #2 53.9153.9427,5281689  0% 1% 1% 100%
3,567Racket 103.92104.0128,5361496  0% 1% 1% 100%
4,127F# Mono 25.0925.1033,0161259  0% 0% 1% 100%
5,360Smalltalk VisualWorks 192.75192.8042,8801652  0% 0% 0% 100%
6,918Hack #3 21 min21 min55,3441080  1% 0% 0% 100%
7,010Hack #5 10 min10 min56,0801432  1% 0% 0% 100%
7,015Hack 13 min13 min56,1161288  1% 0% 0% 100%
7,358Clojure 27.6627.6758,8642162  0% 1% 1% 100%
7,536Clojure #2 27.9027.9260,2841440  1% 0% 0% 100%
86,424Ruby JRuby #2 6 min6 min691,3921137  0% 1% 1% 100%
"wrong" (different) algorithm / less comparable programs
134C++ g++ #2 13.3913.401,0682288
1,191JavaScript V8 #3 32.9132.939,5241505

 n-body benchmark : Double-precision N-body simulation

You can write your own program for this task and contribute to the benchmarks game by following these general instructions.

More specifically:

ndiff -abserr 1.0e-8 program output N = 1000 with this output file to check your program is correct before contributing.

We are trying to show the performance of various programming language implementations - so we ask that contributed programs not only give the correct result, but also use the same algorithm to calculate that result.

Each program should model the orbits of Jovian planets, using the same simple symplectic-integrator - see the Java program.

For background information see N-body problem. Useful symplectic integrators are freely available, for example the HNBody Symplectic Integration Package.

Thanks to Mark C. Lewis for this benchmark.

Revised BSD license

  Home   Conclusions   License   Play