🔢 Primery

A fast prime number generator with multiple algorithms and parallel execution support.

Primery generates prime numbers in a specified range [x, y] and benchmarks the performance of different algorithms. Supports both single-threaded and multi-threaded execution modes.

📊 Supported Algorithms

Algorithm	Alias	Complexity
Trial Division	td	O(n√n)
Sieve of Eratosthenes	soe	O(n log log n)
Sieve of Sundaram	sos	O(n log n)
Sieve of Atkin	soa	O(n)

🚀 Quick Start

Build from Source

# Clone the repository
git clone https://github.com/yourusername/primery.git
cd primery

# Build
make build

# Run
./build/primery soa -i [2,1000000]

Install Globally

sudo make install
primery soa -i [2,1000000]

Using Docker

# Build image
docker build -t primery .

# Run
docker run --rm primery -i [2,100000] -m p soa

📖 Usage

primery [OPTIONS] <algorithm>

Options

Flag	Description	Default
-i, --interval	Range [start,end]	[2,1000]
-m, --mode	s|single or p|parallel	single
-t, --time	ns, ms, or s	ms
-o, --output	Output file path	primes.txt
-h, --help	Show help	-

Examples

# Generate primes from 2 to 1 million using Sieve of Atkin
primery soa -i [2,1000000]

# Parallel execution with nanosecond timing
primery -i [234,100000] -t ns -m p sieveOfEratosthenes

# Save to custom file
primery soe -i [2,10000] -o my_primes.txt

🛠️ Build Commands

make build      # Build the project
make install    # Install to /usr/local/bin
make uninstall  # Remove installation
make clean      # Clean build artifacts
make package    # Create .tar.gz and .deb packages

📈 Benchmarks

Tested on 11th Gen Intel Core i7-11800H @ 2.30GHz × 8

Single-Threaded

n	Eratosthenes	Sundaram	Atkin
10⁵	0.718 ms	0.602 ms	0.936 ms
10⁶	7.544 ms	6.956 ms	8.386 ms
10⁷	66.149 ms	65.015 ms	62.988 ms
10⁸	0.947 s	1.007 s	0.837 s
10⁹	10.956 s	16.559 s	8.839 s

Multi-Threaded

n	Eratosthenes	Sundaram	Atkin
10⁵	3.197 ms	27.105 ms	10.941 ms
10⁶	11.735 ms	140.159 ms	40.063 ms
10⁷	62.772 ms	1.195 s	122.007 ms
10⁸	0.49 s	11.984 s	0.511 s
10⁹	7.284 s	-	3.903 s

💡 Conclusion

The Sieve of Atkin performs best for large ranges (10⁸+) in both single and multi-threaded modes. For smaller ranges, Sieve of Sundaram is competitive in single-threaded mode. Multi-threading provides significant speedup for large inputs, with Atkin achieving ~2.3x improvement at 10⁹.