/mobile Handheld Friendly website

 binary-trees benchmark N=20

Each chart bar shows how many times slower, one ↓ binary-trees 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.

     sortsortsort
  ×   Program Source Code CPU secs Elapsed secs Memory KB Code B ≈ CPU Load
1.0C++ g++ #7 10.3610.3867,008919  0% 1% 0% 100%
1.1Ada 2005 GNAT #4 11.1811.20105,6962167  1% 1% 1% 100%
1.1Ada 2005 GNAT #5 11.2211.25105,6962167  0% 1% 1% 100%
1.5Rust 15.1315.1590,480779  0% 0% 1% 100%
1.6Java  #3 16.2816.30511,448584  1% 1% 0% 100%
1.6Java  #2 16.5616.59510,068603  1% 1% 1% 100%
1.7Scala #4 17.4617.49506,212494  1% 1% 0% 100%
1.8Haskell GHC #4 18.2318.25173,344612  0% 1% 1% 100%
1.8Haskell GHC 18.9118.94166,212521  0% 0% 0% 100%
1.8Lisp SBCL #2 19.1219.15177,812649  0% 1% 0% 100%
2.1C# Mono 21.5921.62120,948654  0% 1% 1% 100%
2.3Fortran Intel #2 23.7723.7999,6161199  0% 0% 0% 100%
2.5F# Mono 25.8425.88189,180537  0% 0% 1% 100%
2.6Racket #2 27.1427.18218,080640  1% 1% 1% 100%
2.8Lisp SBCL 28.5028.54177,784612  0% 0% 1% 100%
2.9Clojure 30.2430.28547,012657  1% 1% 0% 100%
3.0Clojure #2 31.0831.12547,480750  1% 1% 1% 100%
3.2Pascal Free Pascal 33.2533.2765,828769  0% 1% 1% 100%
3.2C gcc 33.4033.4266,104706  0% 0% 0% 100%
3.4OCaml #2 35.0035.07101,028784  0% 0% 0% 100%
3.4Dart 35.0135.04154,676503  0% 1% 0% 100%
3.5OCaml #5 35.8935.93115,712496  0% 0% 0% 100%
3.6F# Mono #3 37.3637.40167,460565  0% 1% 1% 100%
3.6Racket 37.3837.43281,180495  0% 1% 1% 100%
3.7C++ g++ #2 38.0438.0699,092553  0% 1% 1% 100%
3.8JavaScript V8 39.8639.92267,848467  0% 1% 1% 100%
3.9Ada 2005 GNAT 39.9840.0199,444955  1% 1% 0% 100%
4.0Clojure #6 41.0641.11561,116705  1% 1% 0% 100%
4.1C# Mono #2 42.5542.60319,368650  0% 0% 1% 100%
4.6Erlang HiPE #2 47.5747.63338,644499  0% 0% 0% 100%
4.9Erlang HiPE 50.4950.55443,172441  0% 0% 0% 100%
6.3Smalltalk VisualWorks 65.6065.67316,312722  0% 0% 0% 100%
6.5F# Mono #2 67.1467.22281,144515  0% 1% 1% 100%
7.3C gcc #5 75.8275.87109,668963  0% 1% 1% 100%
9.0Go 93.5593.63255,844516  1% 1% 1% 100%
11Go #4 110.33110.42305,600688  0% 1% 0% 100%
11Go #2 112.38112.47305,416694  0% 1% 0% 100%
11Go #5 113.29113.39315,4121000  0% 1% 0% 100%
11Ada 2005 GNAT #3 115.88115.94328,9841342  0% 1% 1% 100%
16Ruby JRuby #3 168.68168.87899,264439  1% 1% 1% 100%
17Fortran Intel 173.18173.35153,800826  0% 0% 0% 100%
21Ruby #2 222.74222.85557,644413  0% 1% 1% 100%
22Ruby #3 226.44226.55409,360439  0% 1% 1% 100%
22Ruby 227.06227.16400,436412  0% 0% 1% 100%
24Ruby JRuby 253.43253.68895,044412  0% 1% 1% 100%
43Lua #2 7 min7 min1,033,624446  0% 14% 4% 100%
49Python 3 #6 8 min8 min688,448626  1% 1% 1% 100%
52PHP #2 9 min9 min546,272472  0% 0% 0% 100%
58Perl 10 min10 min289,324448  0% 0% 0% 100%
60PHP 10 min10 min546,248504  0% 0% 0% 100%
91PHP #3 15 min15 min1,260,280483  1% 1% 4% 100%
C++ g++ #6 Make Error892
Perl #3 Failed706
Racket #3 Bad Output877
Scala #2 Failed641
"wrong" (different) algorithm / less comparable programs
0.3C gcc #2 2.993.0025,152594
0.6C gcc #9 6.236.24114,1281103
0.9C gcc #7 9.479.4899,412850
1.2OCaml 12.7412.77235,724563
1.6Scala 16.1416.16379,632549
1.9Haskell GHC #5 20.1920.2188,332611
3.5Go #6 36.2836.31232,636937
46Lua #3 477.10477.692,963,916477
52PHP #4 541.62542.15957,228945
missing benchmark programs
C CINT No program

 binary-trees benchmark : Allocate and deallocate many many binary trees

diff program output N = 10 with this 1KB 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

Note: this is an adaptation of a benchmark for testing GC so we are interested in the whole tree being allocated before any nodes are GC'd - which probably excludes lazy evaluation.

Note: the left subtrees are heads of the right subtrees, keeping a depth counter in the accessors to avoid duplication is cheating!

Note: the tree should have tree-nodes all the way down, replacing the bottom nodes by some other value is not acceptable; and the bottom nodes should be at depth 0.

Note: these programs are being measured with the default initial heap size - the measurements may be very different with a larger initial heap size or GC tuning.

Please don't implement your own custom memory pool or free list.


The binary-trees benchmark is a simplistic adaptation of Hans Boehm's GCBench, which in turn was adapted from a benchmark by John Ellis and Pete Kovac.

Thanks to Christophe Troestler and Einar Karttunen for help with this benchmark.

Revised BSD license

  Home   Conclusions   License   Play