---

Speed comparison of programming languages

This projects tries to compare the speed of different programming languages. In this project we don't really care about getting a precise calculation of pi. We only want to see how fast are the programming languages doing.

It uses an implementation of the Leibniz formula for π to do the comparison.

Here is a video which explains how it works: Calculating π by hand

Disclaimer

I'm no expert in all these languages, so take my results with a grain of salt.
Also the findings just show how good the languages can handle floating point operations, which is only one aspect of a programming language.

You are also welcome to contribute and help me make fix my possible misuse of some languages. 😄

Languages used in this comparison

• Julia - JIT
• R - interpreted
• Python - interpreted (CPython)
• Ruby - interpreted
• Swift - compiled (in this test interpreted due to Linux Swift limitations)
• Java - compiled, VM
• Rust - compiled
• Javascript using Node.js - interpreted, JIT
• Lua - interpreted
• Nim - compiled
• PHP - interpreted
• C++ - compiled
• Crystal - compiled
• Go - compiled
• C - compiled

Results

======= Comparison =======
Iterations: 10000000

> Julia
Version: julia version 1.1.0
Speed (all): 1253ms, 1287ms, 1300ms, 1282ms, 1342ms, 1300ms, 1285ms, 1257ms, 1290ms, 1314ms
Speed (best): 1253ms
Speed (worst): 1342ms
Speed (median): 1288.5ms
Result: 

> Python 3 (CPython)
Version: Python 3.7.3
Speed (all): 2663ms, 2602ms, 2663ms, 2637ms, 2590ms, 2708ms, 2562ms, 2667ms, 2517ms, 2504ms
Speed (best): 2504ms
Speed (worst): 2708ms
Speed (median): 2619.5ms
Result: 3.1415927535897814

> R
Version: R version 3.6.0 (2019-04-26) -- "Planting of a Tree"
Speed (all): 1893ms, 1876ms, 1876ms, 1879ms, 1911ms, 1886ms, 1886ms, 1894ms, 1876ms, 1898ms
Speed (best): 1876ms
Speed (worst): 1911ms
Speed (median): 1886.0ms
Result: 3.1415925535898115

> Ruby
Version: ruby 2.6.3p62 (2019-04-16 revision 67580) [x86_64-linux]
Speed (all): 1229ms, 1236ms, 1225ms, 1328ms, 1222ms, 1211ms, 1277ms, 1228ms, 1371ms, 1382ms
Speed (best): 1211ms
Speed (worst): 1382ms
Speed (median): 1232.5ms
Result: 3.1415925535898115

> Java
Version: 8
Speed (all): 145ms, 138ms, 130ms, 123ms, 130ms, 127ms, 124ms, 131ms, 137ms, 134ms
Speed (best): 123ms
Speed (worst): 145ms
Speed (median): 130.5ms
Result: 3.1415927535897814

> Lua
Version: Lua 5.3.5  Copyright (C) 1994-2018 Lua.org, PUC-Rio
Speed (all): 801ms, 871ms, 828ms, 777ms, 870ms, 816ms, 777ms, 853ms, 832ms, 863ms
Speed (best): 777ms
Speed (worst): 871ms
Speed (median): 830.0ms
Result: 3.1415925535898115

> Rust
Version: rustc 1.34.1
Speed (all): 432ms, 426ms, 428ms, 425ms, 428ms, 424ms, 425ms, 425ms, 424ms, 427ms
Speed (best): 424ms
Speed (worst): 432ms
Speed (median): 425.5ms
Result: 3.1415927535897814

> JS (node)
Version: v11.15.0
Speed (all): 123ms, 115ms, 116ms, 114ms, 117ms, 114ms, 120ms, 128ms, 123ms, 113ms
Speed (best): 113ms
Speed (worst): 128ms
Speed (median): 116.5ms
Result: 3.1415927535897814

> PHP 7
Version: PHP 7.3.5 (cli) (built: Apr 30 2019 21:05:09) ( NTS )
Speed (all): 452ms, 447ms, 447ms, 448ms, 446ms, 443ms, 456ms, 473ms, 448ms, 444ms
Speed (best): 443ms
Speed (worst): 473ms
Speed (median): 447.5ms
Result: 3.1415927535897814

> Python 3 (pypy)
Version: 7.1.1-1
Speed (all): 255ms, 158ms, 157ms, 152ms, 161ms, 154ms, 167ms, 161ms, 153ms, 151ms
Speed (best): 151ms
Speed (worst): 255ms
Speed (median): 157.5ms
Result: 3.1415927535897814

> Nim
Version: 0.18.0-1
Speed (all): 481ms, 453ms, 430ms, 427ms, 421ms, 423ms, 425ms, 424ms, 426ms, 423ms
Speed (best): 421ms
Speed (worst): 481ms
Speed (median): 425.5ms
Result: 3.1415925535898115

> C++
Version: g++ (GCC) 8.3.0
Speed (all): 42ms, 35ms, 34ms, 34ms, 37ms, 35ms, 34ms, 34ms, 38ms, 35ms
Speed (best): 34ms
Speed (worst): 42ms
Speed (median): 35.0ms
Result: 3.1415927535897814

> Crystal
Version: Crystal 0.28.0 (2019-04-17)
Speed (all): 59ms, 38ms, 42ms, 37ms, 39ms, 38ms, 36ms, 38ms, 40ms, 37ms
Speed (best): 36ms
Speed (worst): 59ms
Speed (median): 38.0ms
Result: 3.1415925535898115

> C
Version: gcc (GCC) 8.3.0
Speed (all): 37ms, 39ms, 35ms, 33ms, 33ms, 33ms, 33ms, 33ms, 34ms, 33ms
Speed (best): 33ms
Speed (worst): 39ms
Speed (median): 33.0ms
Result: 3.1415927535897814

> Go
Version: go version go1.12.4 linux/amd64
Speed (all): 41ms, 19ms, 18ms, 18ms, 19ms, 18ms, 19ms, 18ms, 17ms, 17ms
Speed (best): 17ms
Speed (worst): 41ms
Speed (median): 18.0ms
Result: 3.1415927535897814

Run it yourself

Everything is run by a Docker container and a bash script which envokes the programs.

To measure the execution time a python package is used.

Requirements

• Docker
• Makefile support

Run measurement

Just run: make

Print plot

make plot

Run cli

make cli

FAQ

Why do you also count reading a file and printing the output?

Because I think this is a more realistic scenario to compare speeds.

Are the compile times included in the measurements?

No they are not included, because when running the program in the real world this would also be done before.

TODO

• Add C++
• Add run pypy for Python
• Produce graphic with results
• Add Java
• Add Swift
• Add R
• Add Php 5.6
• Add phpv8js
• Add python2
• Add Smalltalk
• Add C#
• Add Perl
• Add Kotlin
• Add Haskell
• Add Elixir
• Add Lisp
• Add Ada
• Add COBOL

Thanks

This projects takes inspiration from Thomas who did a similar comparison on his blog.