Recursive Fibonacci Benchmark using top languages on Github

Top 10: JavaScript, Java, Python, Ruby, Php, C++, C#, C, Go reference

Others: Crystal, Rust, Swift, Mono, Elixir, Perl, R, Julia, D, Nim

This code performs a recursive fibonacci to the 46th position with the result of 2,971,215,073.

Fibonacci can be written many different ways. The goal of this project is to compare how each language handles the exact same code.

Here is the Crystal version:

def fib(n : UInt64)
  return 1_u64 if n <= 1
  fib(n - 1) + fib(n - 2)
end

puts fib(46)

Here is the Ruby version:

def fib(n)
  return 1 if n <= 1
  fib(n - 1) + fib(n - 2)
end

puts fib(46)

All tests are run on:

iMac (Retina 5K, 27-inch, Late 2015)
OS: macOS High Sierra 10.13.6
Processor: 3.2 GHz Intel Core i5
Memory: 16 GB 1867 MHz DDR3

Last benchmark was ran on September 29, 2018

Natively compiled, statically typed

Language	Time, s	Compile	Run
Nim	4.678	nim cpp -d:release fib.nim	time ./fib
C	5.418	gcc -O3 -o fib fib.c	time ./fib
C++	5.456	g++ -O3 -o fib fib.cpp	time ./fib
Crystal	5.800	crystal build --release fib.cr	time ./fib
Fortran	6.139	gfortran -O3 -o fib fib.f03	time ./fib
Rust	6.724	rustc -O fib.rs	time ./fib
D	7.124	ldc2 -O3 -release -flto=full -of=fib fib.d	time ./fib
OCaml	8.131	ocamlopt -O3 -o fib fib.ml	time ./fib
Haskell	8.143	ghc -O3 -o fib fib.hs	time ./fib
Swift	10.550	swiftc -O -g fib.swift	time ./fib
Go	10.863	go build fib.go	time ./fib
Cython		`cython3 --embed -o fib.pyx.c fib.pyx &&` `gcc -O3 -o fib fib.pyx.c $(pkg-config --cflags --libs python3)`	`time ./fib`

NOTE: C and C++ compile to the exact same machine code but the C version is slower because it doesn't run at the same address in the processor. Thanks @glandium for pointing this out. See Issue #46

VM compiled bytecode, statically typed

Language	Time, s	Compile	Run
Java	7.556	javac Fib.java	time java Fib
C#	11.387	dotnet build -c Release -o ./bin	time dotnet ./bin/fib.dll
C# (Mono)	12.310	mcs fib.cs	time mono fib.exe

VM compiled, dynamically typed

Language	Time, s	Compile	Run
Erlang		`erlc +native +'{hipe,[o3]}' fib.erl`	`time erl -noimput -noshell -s fib`

VM compiled before execution, mixed/dynamically typed

Language	Time, s	Run
Dart	9.651	time dart fib.dart
Julia	11.461	time julia -O3 fib.jl
Elixir*	13.955	time elixir fib.exs
Node	19.161	time node fib.js
Escript(Erlang)		`time escript fib.es`

Elixir is using ERL_COMPILER_OPTIONS='[native,{hipe, [o3]}]'

NOTE: These languages include compilation time that should be taken into consideration when comparing.

Interpreted, dynamically typed

Language	Time, s	Run
Php	198.279	time php fib.php
Ruby	202.901	time ruby fib.rb
Scheme	???????	`time guile fib.scm`
Python	512.621	time python fib.py
Python3	758.681	`time python3 fib.py`
Perl	1133.131	`time perl fib.pl`
Perl 6	TODO	`time perl6 fib.p6`
Tcl	TODO	`time tclsh fib.tcl`
Lua	TODO	`time lua fib.lua`
R	1796.495	`time r -f fib.r`
K	TODO	`time k fib.k`

NOTE: Interpreted languages have a startup time cost that should be taken into consideration when comparing.

Optimized code that breaks the benchmark

The code examples provided in the optimized folder use techniques that break the benchmark. They do not perform the same internal tasks as the other examples so are not a good for an apples to apples comparison. They all perform at sub-second response times. It demonstrates that all benchmarks will have some caveat.

Several of these examples add very little changes to the original code:

The C++ constexpr is using a constexpr which optimizes the recursive call to a constant.
The Python lru_cache is using lru_cache directive with no code changes.
The Ruby mem and JS mem are maintaining a simple cache in an array.

Versions

go version go1.11 darwin/amd64
g++-8 (Homebrew GCC 8.2.0) 8.2.0
crystal Crystal 0.26.1 (2018-08-27) LLVM: 6.0.1
g++ Apple LLVM version 10.0.0 (clang-1000.11.45.2)
gcc Apple LLVM version 10.0.0 (clang-1000.11.45.2)
nim Nim Compiler Version 0.18.0 [MacOSX: amd64]
swiftc Apple Swift version 4.2 (swiftlang-1000.11.37.1 clang-1000.11.45.1)
rustc 1.29.0
javac 10.0.1
mcs Mono C# compiler version 5.12.0.226
dotnet 2.1.4
dart Dart VM version: 2.0.0 (Fri Aug 3 10:53:23 2018 +0200)
julia version 0.6.3
node v9.4.0
elixir Elixir 1.7.3 (compiled with Erlang/OTP 21)
ruby 2.5.1p57 (2018-03-29 revision 63029)
php 7.1.16 (cli) (built: Apr 1 2018 13:14:42)
python 2.7.15
python3 3.7.0
perl 5, version 26, subversion 2 (v5.26.2)
perl6 This is Rakudo Star version 2018.06 built on MoarVM version 2018.06
r version 3.5.0 (2018-04-23)
lcd2 the LLVM D compiler (1.11.0)
ocaml The OCaml toplevel, version 4.07.0
ghc The Glorious Glasgow Haskell Compilation System, version 8.4.3
gfortran GNU Fortran (Homebrew GCC 8.2.0) 8.2.0
tchsh 8.5

Caveats

Fibonacci Benchmark

CharlesHD/fib