ali77gh / language_performance_prime_algorithm

implementing is_prime function and test speed of execution in different languages and runtimes.

Performance Test

This repository contains an implementation of the is_prime function in multiple programming languages using the same algorithm. The goal is to compare the performance of different languages and runtimes.

Usage

To run code in various programming languages, use the run.sh script:

./run.sh [LANGUAGE]

If no argument is provided, the script will run code in all supported languages. Otherwise, provide a language name to run the corresponding code.

Supported Languages

• assembly
• c
• cpp
• rust
• go
• java
• nodejs
• csharp_mono
• dart
• python_codon
• pascal
• python
• php
• r
• chap
• lua

Examples

# Run code in all supported languages
./run.sh

# Run specific language
./run.sh cpp

To display usage instructions and a list of supported languages, use the --help or -h flag:

./run.sh --help

If you want to add your own test results, simply run the run.sh script with your CPU model name and redirect the output to a file in the result directory.

See the result directory for the output of running the tests on different CPUs.

My Laptop (Intel(R) Core(TM) i7-8550U)

Note: times are 5% more or less on different runs, so C and C++ can win sometimes.

Ranking	Language	Runtime/Compiler	Time	n% Slower than Winner
#1	Assembly	NASM	3.10s	Winner
#2	C	GCC	3.12s	0.64%
#3	C++	G++	3.15s	1.6%
#4	Java	OpenJDK 17	3.62s	16%
#5	Go	Go 1.20.3	3.81s	22%
#6	Julia	Julia 1.8.5	3.87s	24%
#7	Rust	Native 1.67.1	3.97s	28%
#8	C#	Mono 6.8	5.32s	71%
#9	Js	Nodejs 20	5.80s	87%
#10	Dart	Dart 2.19.0	7.11s	129%
#11	Python	Codon 0.16.0	10.89s	251%
#12	Pascal	FPC 3.2.2	13.68s	341%
#13	PHP	PHP 8.2.5	26.32s	749%
#14	Lua	Lua 5.2.4	34.00s	996%
#15	Python	Python 3.11.3	82.31s	2545%
#16	Ruby	Ruby 3.2.2	85.55s	2641%
#17	R	Rscript 4.0.4	240.0s	7641%
#18	Chap	chap 1.0.1	1025s	32964%

The n% Slower than Winner column shows the percent difference between the time taken by each language and the fastest language (Assembly).

Formula: ((LanguageTime - WinnerTime) * 100) / WinnerTime

Contribute

You can contribute to this repository in one of the following ways:

1. Optimize your favorite programming language without changing the algorithm and without using concurrency.
2. Add a new language by using the same algorithm.
3. Add a new runtime.

language_performance_prime_algorithm's Issues

Add Contrbutors to README

Github and Twitter accounts.
Also say thanks to rejected PRs on Twitter and README.

Update some runtimes

Happy birthday : )

C/C++ must be compiled with -O2 not -O3

-O2 has better performance usually, I'm told.

Related: Benchmarking Common Lisp and C++ to find prime numbers

Sometime ago, I've done a similar benchmark. You can use the code from there in your benchmark.

https://far.chickenkiller.com/computer/cpp-vs-common-lisp-performance/

Measure timings with unix time utility not language utilities

Hello. It is obvious the at when you measure time within the language itself, the "fire up" time will be eliminated from the result leaving an incorrect result.

add cython

hi can i add cython version of python?