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++ #8 9.089.083361544  0% 0% 1% 100%
1.0C++ g++ #7 9.109.103361545  2% 0% 5% 100%
1.1C gcc #4 9.919.923361490  0% 1% 1% 100%
1.1C++ g++ #3 9.949.956201763  1% 0% 1% 100%
1.2Fortran Intel #5 10.7410.745201659  1% 1% 1% 100%
1.4C++ g++ #5 12.7412.758681749  0% 1% 1% 100%
1.7Ada 2005 GNAT #5 15.0815.091,2122427  1% 0% 0% 100%
2.2C++ g++ 20.2520.263361659  0% 1% 0% 100%
2.2C gcc 20.3220.323361173  0% 0% 1% 100%
2.3C++ g++ #6 20.4820.493361668  1% 1% 0% 100%
2.3Fortran Intel #2 20.5320.545241496  0% 0% 0% 100%
2.3C gcc #6 20.6820.693361180  0% 1% 1% 100%
2.3Fortran Intel 20.8820.895201389  1% 1% 0% 100%
2.3C gcc #3 21.0321.033361208  0% 0% 1% 100%
2.4C# Mono #3 22.0422.0419,4121305  0% 1% 1% 100%
2.4C++ g++ #4 22.2422.252841428  1% 1% 0% 100%
2.5C# Mono #8 22.4122.4219,4161343  1% 0% 1% 100%
2.5C gcc #5 22.5722.582841429  0% 1% 0% 100%
2.5Fortran Intel #4 22.6022.605201172  1% 0% 0% 100%
2.5C gcc #2 22.7022.713361263  0% 0% 1% 100%
2.5C# Mono #2 23.0923.1019,1921410  0% 1% 1% 100%
2.5Go 23.0923.107641310  0% 1% 0% 100%
2.6Ada 2005 GNAT #3 23.4623.471,2121740  1% 0% 0% 100%
2.6Scala 23.8423.8525,0761358  1% 1% 0% 100%
2.7Java  24.4824.4919,3081424  1% 1% 1% 100%
2.7Java  #2 24.5024.5219,6361424  1% 0% 1% 100%
2.7Rust 24.5824.594,8521474  0% 1% 1% 100%
2.7C# Mono #6 24.6724.6819,2881289  1% 1% 0% 100%
2.8Fortran Intel #3 25.2125.225161299  1% 0% 0% 100%
2.8F# Mono 25.6625.6727,1681259  0% 1% 1% 100%
2.9Ada 2005 GNAT 26.0726.081,2161608  1% 1% 0% 100%
2.9Haskell GHC #2 26.0726.082,2121874  1% 0% 0% 100%
3.1Haskell GHC 28.5528.562,2081706  0% 1% 0% 100%
3.2Clojure 29.1629.1748,5122162  0% 1% 1% 100%
3.7Dart #2 33.9233.9417,8121274  0% 1% 1% 100%
3.8Pascal Free Pascal 34.7134.7281308  0% 0% 0% 100%
4.1Pascal Free Pascal #2 37.2037.2281322  1% 0% 0% 100%
4.2Clojure #2 37.8437.8648,3881440  1% 0% 1% 100%
4.2Lisp SBCL #2 38.4238.437,8521398  1% 0% 0% 100%
4.2C# Mono 38.5038.5219,6241403  0% 1% 1% 100%
4.6JavaScript V8 #2 41.6541.666,8681527  0% 5% 1% 100%
4.9JavaScript V8 44.6644.676,9681287  0% 0% 1% 100%
5.1Lisp SBCL 46.0246.037,8521363  0% 0% 1% 100%
5.2Dart #3 46.8146.8318,1121420  0% 1% 1% 100%
5.2Pascal Free Pascal #3 47.1847.1981418  0% 0% 1% 100%
6.1Racket #2 55.5255.5422,1681689  1% 1% 1% 100%
8.3OCaml 75.2175.236001239  1% 0% 1% 100%
13Racket 116.12116.2035,6601496  0% 1% 1% 100%
24Erlang HiPE #3 217.95218.0210,4601399  1% 0% 0% 100%
35Smalltalk VisualWorks 5 min5 min23,4521652  0% 0% 1% 100%
43Ruby JRuby #2 6 min6 min713,5801137  0% 1% 1% 100%
65Lua #4 9 min9 min8721305  0% 0% 1% 100%
66Lua #2 9 min9 min8721193  0% 1% 1% 100%
78Lua 11 min11 min8721201  0% 1% 1% 100%
79PHP #3 11 min11 min2,5841082  0% 0% 1% 100%
109Python 3 16 min16 min4,2321181  1% 0% 1% 100%
128Perl #2 19 min19 min2,2641401  0% 1% 1% 100%
162Ruby #2 24 min24 min5,1241137  0% 1% 1% 100%
7,973C CINT 20h 06 min20h 06 min5,2881173  0% 0% 0% 100%
C# Mono #5 Failed1485
C# Mono #7 Failed1539
C# Mono #4 Failed1461
"wrong" (different) algorithm / less comparable programs
1.3C++ g++ #2 12.1112.118682288
3.6JavaScript V8 #3 32.9832.998,7321505

 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