https://github.com/vladimir-sama/pakhomov_gsc
Pakhomov GSC (Generative Seed Compression)
https://github.com/vladimir-sama/pakhomov_gsc
compression compression-algorithm cpp multithreading pseudo-random-generator pseudorandom
Last synced: 6 months ago
JSON representation
Pakhomov GSC (Generative Seed Compression)
- Host: GitHub
- URL: https://github.com/vladimir-sama/pakhomov_gsc
- Owner: vladimir-sama
- License: gpl-3.0
- Created: 2025-04-22T06:00:27.000Z (6 months ago)
- Default Branch: main
- Last Pushed: 2025-04-22T06:12:20.000Z (6 months ago)
- Last Synced: 2025-04-22T07:39:50.092Z (6 months ago)
- Topics: compression, compression-algorithm, cpp, multithreading, pseudo-random-generator, pseudorandom
- Language: C++
- Homepage:
- Size: 19.5 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Pakhomov GSC
Vladimir Alexandre Pakhomov
Concept and Implementation: 2025
**Pakhomov GSC** (Generative Seed Compression) is an experimental file compression concept based on deterministic pseudo-random generation.
The idea is simple: for any file, there exists a seed and a length such that a deterministic pseudo-random number generator (PRNG) will generate the exact bit sequence of the original file. If such a seed and length are found, the file can be represented by just those two numbers.
This project explores that concept in C++ with a fast multithreaded implementation.
## How It Works
1. The program reads a binary file.
2. It converts the file to a vector of bits.
3. Using a seeded PRNG (e.g., Xorshift), it attempts to generate a matching bit sequence of the same length.
4. It searches through seeds in parallel using multiple CPU threads.
5. If a match is found, it outputs:
- The matching seed
- The bit length
This effectively "compresses" the file to two numbers: the PRNG seed and the bit length. While practically unfeasible for large files, this is a conceptual exploration of ultimate compression and entropy mapping.
## Use Case
This project is intended for research, theoretical exploration, and experimentation in data entropy, compression theory, and algorithmic generation. It is not a practical compression tool.
## Build Instructions
```bash
g++ -std=c++20 -O3 -pthread -o pakhomov_gsc pakhomov_gsc.cpp
```