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
1.0Fortran Intel #5 8.578.582601659  1% 1% 100% 1%
1.1C++ g++ #8 9.089.093361544  0% 100% 1% 0%
1.1C++ g++ #7 9.109.103361545  0% 0% 1% 100%
1.2C gcc #4 9.919.923361490  0% 0% 1% 100%
1.2C++ g++ #3 9.949.956201763  1% 100% 0% 0%
1.5C++ g++ #5 12.7612.768681749  0% 1% 1% 100%
1.8Ada 2005 GNAT #5 15.0815.091,2122427  1% 0% 0% 100%
2.3Fortran Intel #2 20.0520.065201496  0% 0% 0% 100%
2.4C++ g++ 20.2620.263361659  0% 100% 1% 1%
2.4C gcc 20.3220.333361173  0% 100% 0% 1%
2.4Fortran Intel 20.3420.352601389  0% 0% 0% 100%
2.4C++ g++ #6 20.4820.493361668  0% 0% 1% 100%
2.4C gcc #6 20.6720.683361180  0% 0% 1% 100%
2.5C gcc #3 21.0321.043361208  0% 0% 1% 100%
2.6Rust 21.9021.914,6801371  1% 100% 0% 1%
2.6C# Mono #3 22.0422.0519,4081305  100% 1% 1% 1%
2.6C++ g++ #4 22.2422.252841428  0% 100% 0% 1%
2.6C gcc #5 22.5822.582841429  1% 100% 0% 1%
2.6C gcc #2 22.7022.713361263  0% 0% 1% 100%
2.7Fortran Intel #4 22.8222.832521172  0% 0% 100% 0%
2.7Fortran Intel #3 22.8922.902561299  0% 0% 0% 100%
2.7Rust #3 22.9722.984,6841367  0% 1% 0% 100%
2.7C# Mono #2 23.0923.1019,1921410  1% 0% 100% 1%
2.7Go 23.1423.137601310  0% 1% 100% 1%
2.7Scala 23.4923.4322,8521358  1% 1% 1% 100%
2.7Ada 2005 GNAT #3 23.4723.481,2161740  1% 0% 1% 100%
2.8Haskell GHC #2 25.0024.402,1921874  1% 1% 100% 1%
2.8Rust #2 24.4324.444,6801321  0% 100% 1% 0%
2.9Java  #2 24.4724.4519,5081424  1% 1% 100% 0%
2.9C# Mono #6 24.6724.6819,2881289  100% 0% 1% 1%
3.0F# Mono 25.6725.6827,1761259  1% 100% 0% 1%
3.0Ada 2005 GNAT 26.0726.081,2161608  1% 1% 0% 100%
3.3Clojure 29.4228.0349,3922162  4% 2% 97% 3%
3.4Haskell GHC 30.9329.132,2121706  100% 2% 2% 2%
4.0Dart #2 34.7734.5312,9921274  1% 1% 100% 1%
4.0Pascal Free Pascal 34.6934.7081308  1% 0% 100% 0%
4.3Clojure #2 38.0136.7148,9641440  3% 95% 3% 5%
4.3Pascal Free Pascal #2 37.2037.2281322  0% 1% 1% 100%
4.5Lisp SBCL #2 38.2438.258,0081398  1% 0% 0% 100%
4.5C# Mono 38.4938.5119,6241403  1% 100% 0% 0%
4.5Dart #3 39.0238.7413,2721420  1% 1% 100% 1%
5.5Lisp SBCL 47.1447.158,0081363  1% 1% 1% 100%
5.5Pascal Free Pascal #3 47.1847.1981418  0% 1% 1% 100%
5.8OCaml 49.9249.935961239  0% 0% 0% 100%
6.5Racket #2 56.0456.0417,6641689  0% 0% 1% 100%
14Racket 116.05116.0620,7881496  0% 0% 0% 100%
26Erlang HiPE #3 225.14225.2112,9481399  100% 1% 1% 1%
45Ruby JRuby #2 6 min6 min711,9601137  13% 32% 40% 19%
86PHP #3 12 min12 min2,5961082  0% 38% 62% 0%
115Python 3 16 min16 min4,2521181  0% 100% 0% 0%
137Perl #2 19 min19 min2,2801401  24% 16% 61% 1%
185Ruby #2 26 min26 min5,4841137  0% 1% 38% 63%
C# Mono #5 Failed1485
C# Mono #7 Failed1539
C# Mono #4 Failed1461
"wrong" (different) algorithm / less comparable programs
1.4C++ g++ #2 11.9311.938682288

 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