Create your own Ranking

What fun! Can you manipulate the multipliers and weights to make your favourite language the best programming language in the Benchmarks Game?


  x   language mean missing
1.0C++ Intel 1.203
1.2Intel 1.382
1.2GNU gcc 1.442
1.3C++ GNU g++ 1.55 
1.3Lisaac 1.554
1.4Clean 1.653
1.4Pascal Free Pascal 1.682
1.5Digital Mars 1.75 
1.7Oberon-2 OO2C 1.997
1.7Java 6 -Xms64m 2.00 
1.7Fortran Intel 2.065
1.7Haskell GHC 2.081
1.8Eiffel SmartEiffel 2.113
1.8Java 6 -server 2.13 
2.0Lisp SBCL 2.383
2.0Java 1.4 -server 2.382
2.0OCaml 2.384
2.0BASIC FreeBASIC 2.402
2.0Ada 2005 GNAT 2.432
2.0SML MLton 2.453
2.2Java 6 -client 2.61 
2.3Scala 2.742
2.6Java GNU gcj 3.123
2.6Nice 3.161
2.7C# Mono 3.202
2.7CAL 3.231
3.4Forth bigForth 4.053
3.6Fortran G95 4.336
4.9Mercury 5.817
5.4Lua LuaJIT 6.483
5.9Scheme Ikarus 7.075
5.9Python Psyco 7.10 
6.7Erlang HiPE 8.001
7.3Scheme PLT 8.69 
9.5Smalltalk VisualWorks 11.311
11Forth GNU GForth 13.673
12Java 6 -Xint 14.71 
12Scheme Chicken 14.784
16Lua 19.343
17Pike 20.094
17Mozart/Oz 20.602
19Python 23.00 
24Perl 28.273
25PHP 30.003
28Ruby Core 1.9.0 33.565
31Groovy 37.223
34Python IronPython 41.184
39Smalltalk GNU 47.125
43Smalltalk Squeak 51.926
45Ruby JRuby 53.351
48Tcl 57.443
55JavaScript SpiderMonkey 66.286
63Ruby 75.121
multipliers
Time secs
Memory KB
Size B
benchmarkweight
binary-trees

chameneos-redux

fannkuch

fasta

k-nucleotide

mandelbrot

meteor-contest

n-body

nsieve

nsieve-bits

partial-sums

pidigits

recursive

regex-dna

reverse-complement

spectral-norm

startup

sum-file

thread-ring

 about the Ranking

(Language implementations with more than a couple of Timeouts distort the ranking - so they have been excluded. They can still be compared directly against another language implementation.)

For each benchmark, the best measurement B is the lowest non-zero measurement. For each language implementation, the measurement L for the language implementation is converted to the ratio L/B

The ratio can be adjusted by a measurement multiplier (make CPU time or memory use or lines-of-code more or less important) and by a benchmark weight (make particular benchmarks more or less important).

For each language implementation, we report the Weighted Geometric Mean of those L/B ratios.

Revised BSD license