n-body benchmark N=50,000,000

Each chart bar shows how many times slower, one ↓ n-body program was, compared to the fastest program.

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.

    sort sortsort
  ×   Program Source Code CPU secs Elapsed secs Memory KB Code B ≈ CPU Load
1.0C++ g++ #7 9.129.138761545  0% 0% 100% 1%
1.0C++ g++ #8 9.139.147961544  1% 100% 0% 0%
1.1C gcc #4 10.3810.397361490  0% 100% 1% 1%
1.2Fortran Intel #5 10.7510.765201659  1% 0% 100% 1%
1.4C++ g++ #5 12.4812.491,2081749  100% 0% 0% 1%
2.2C++ g++ #3 19.8719.881,2281763  1% 0% 0% 100%
2.2C++ g++ #6 20.1120.128081668  1% 2% 1% 100%
2.2C gcc 20.1920.197161173  0% 0% 100% 0%
2.2C++ g++ #4 20.3020.316281428  1% 0% 0% 100%
2.2Fortran Intel #2 20.5320.545241496  0% 1% 1% 100%
2.3C++ g++ 20.5720.588041659  0% 100% 1% 0%
2.3C gcc #6 20.5920.607881180  1% 0% 100% 0%
2.3Fortran Intel 20.9020.915201389  1% 0% 0% 100%
2.3C gcc #5 21.0721.085801429  1% 0% 0% 100%
2.3C gcc #3 21.3021.317841208  1% 100% 0% 0%
2.5Fortran Intel #4 22.6022.615201172  1% 0% 100% 1%
2.5Go 23.1423.137601310  0% 1% 100% 1%
2.5C# Mono #2 23.1623.1738,7521410  1% 0% 0% 100%
2.6Scala 23.4523.4028,8321358  1% 1% 100% 1%
2.7Rust 24.5724.586,2361474  0% 100% 1% 0%
2.7Ada 2005 GNAT 24.8224.831,7241608  1% 0% 0% 100%
2.8Fortran Intel #3 25.2225.235161299  0% 1% 100% 0%
2.8C gcc #2 25.2725.287281263  100% 0% 1% 0%
2.8Ada 2005 GNAT #3 25.5925.601,7121740  0% 0% 100% 0%
2.8F# Mono 25.9926.0046,8601259  1% 0% 100% 1%
2.9Haskell GHC #2 27.1026.373,6441874  100% 2% 1% 2%
3.1Clojure 29.4228.0349,3922162  4% 2% 97% 3%
3.2Haskell GHC 30.9329.1712,4001706  2% 100% 2% 2%
3.7Dart #2 33.9233.7018,4641274  100% 1% 1% 1%
3.8Pascal Free Pascal 34.6934.7081308  1% 0% 100% 0%
4.0Clojure #2 38.0136.7148,9641440  3% 95% 3% 5%
4.1Pascal Free Pascal #2 37.2037.2281322  0% 1% 1% 100%
4.2Lisp SBCL #2 38.4138.437,8521398  0% 1% 100% 0%
4.2C# Mono 38.5838.5939,4241403  100% 0% 0% 1%
5.0Lisp SBCL 46.0246.037,8521363  0% 0% 100% 0%
5.1Dart #3 46.9446.5518,6521420  0% 1% 1% 100%
5.2Pascal Free Pascal #3 47.1847.1981418  0% 1% 1% 100%
6.1Racket #2 55.3255.3320,9281689  1% 43% 58% 0%
8.2OCaml 75.2375.256001239  1% 0% 0% 100%
13Racket 116.11116.1036,5361496  0% 100% 1% 1%
25Erlang HiPE #3 225.14225.2112,9481399  100% 1% 1% 1%
51Ruby JRuby #2 7 min7 min655,2681137  36% 21% 24% 23%
74PHP #3 11 min11 min2,5921082  0% 1% 1% 100%
108Python 3 #2 16 min16 min6,0801228  6% 100% 2% 7%
109Python 3 16 min16 min4,2401181  0% 0% 100% 1%
135Perl #2 20 min20 min2,5521401  1% 0% 0% 100%
163Ruby #2 24 min24 min5,0841137  22% 0% 79% 1%
Ada 2005 GNAT #5 Make Error2427
C# Mono #8 Failed1343
C# Mono #5 Failed1485
C# Mono #3 Failed1305
C# Mono #7 Failed1539
C# Mono #6 Failed1289
C# Mono #4 Failed1461
Java  #2 Failed1424
Java  Failed1424
"wrong" (different) algorithm / less comparable programs
1.3C++ g++ #2 12.2912.301,2082288

 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