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.

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 25th, 2018

Natively compiled, statically typed

Language	Time, s	Compile	Run
Nim	4.622	`nim cpp -d:release fib.nim`	`time ./fib`
Crystal	5.687	`crystal build --release fib.cr`	`time ./fib`
C++	5.751	`g++ -O3 -o fib fib.cpp`	`time ./fib`
C	6.258	`gcc -O3 -o fib fib.c`	`time ./fib`
Rust	6.567	`rustc -O fib.rs`	`time ./fib`
D	6.993	`ldc2 -O3 -release -flto=full -of=fib fib.d`	`time ./fib`
Swift	10.307	`swiftc -O -g fib.swift`	`time ./fib`
Go	10.600	`go build fib.go`	`time ./fib`
OCaml		`ocamlopt -O3 -o fib fib.ml`	`time ./fib`
Haskell		`ghc -O3 -o fib fib.hs`	`time ./fib`
Fortran	x.xxx	`gfortran -O3 -o fib fib.f03`	`time ./fib`

VM compiled bytecode, statically typed

Language	Time, s	Compile	Run
Java	7.447	`javac Fib.java`	`time java Fib`
C#	7.874	`dotnet build -c Release -o ./bin`	`time dotnet ./bin/fib.dll`
C# (Mono)	12.596	`mcs fib.cs`	`time mono fib.exe`

VM compiled before execution, mixed/dynamically typed

Language	Time, s	Run
Dart	10.467	`time dart fib.dart`
Julia	10.799	`time julia fib.jl`
Node	18.874	`time node fib.js`
Elixir	69.101	`time elixir fib.exs`

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

Interpreted, dynamically typed

Language	Time, s	Run
Ruby	195.601	`time ruby fib.rb`
Php	206.346	`time php fib.php`
Python	502.036	`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`
R	1796.495	`time r -f fib.r`
Tcl	TODO	`time tclsh fib.tcl`

Optimized code that breaks the benchmark

The following code examples use techniques that break the benchmark. They do not perform the same internal tasks as the other examples so are not a good apples to apples comparisons. It demonstrates that all benchmarks will have some caveat.

Language	Time, s	Compile	Run
Go (mem)	0.005*	`go build -o fib fib-mem.go`	`time ./fib`
Nim (mem)	0.006*	`nim cpp -d:release fib_mem.nim`	`time ./fib_mem`
C++ (constexpr)	0.086*	`g++-8 -O3 -o fib fib-constexpr.cpp`	`time ./fib`
Node (mem)	0.112*		`time node fib-mem.js`
Python (lru_cache)	TODO		`time python3 fib-cache.pu`
Lua (mem)	TODO		`time luajit fib-mem.lua`

NOTE: The C++ (constexpr) is using a constexpr which optimizes the recursive call to a constant. It was provided by Ole Christian Eidheim. The Go (mem) is using memoization. It was provided by Alexander F. Rødseth. The Node (mem) is another example using memoization. It was provided by YSTYLE-L.X.Y The Nim (mem) version is provided by PMunch

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)
r version 3.5.0 (2018-04-23)
lcd2 the LLVM D compiler (1.11.0)

Caveats

Fibonacci Benchmark

ntrepid8/fib