Argent Flux is a multi-physics simulation program built to explore how physical and intelligent systems interact — from heat transfer and fluid motion to machine-learning-driven control.
It is developed under the Argent Letters brand, where each Greek letter represents a distinct line of products, ideas, and experiments.

Flux aims to be a unified foundation:
a framework capable of simulating and learning from any system that can be described by equations, data, or both.

What it Can Do

Flux solves the equations that describe how things change — temperature, velocity, concentration, or any other field that evolves in space and time.
It supports 1D, 2D, and 3D domains, and is being designed to model:

Heat transfer in electrical and mechanical systems
Fluid and turbulence dynamics with machine-learning-aided closure models
Control and decision systems, such as drones or robotic agents acting within simulated environments
Data-driven hybrid problems where experimental or AI models interact with physical laws

Why Build It

The project grew from the collaboration between engineers and researchers with backgrounds in power systems, fluids, and data-driven modeling.
The goal is not to reinvent existing solvers, but to merge classical physics with modern computation to give engineers and scientists a single tool that can model, experiment, and learn.

Flux is written in modern C++, using:

OpenMP and MPI for parallel and cluster execution
Planned GPU acceleration for large-scale problems
A design philosophy focused on clarity, reproducibility, and extendibility

Ultimately, we want Flux to scale from a laptop test run to a full HPC cluster, making high-fidelity simulation and machine learning accessible in the same environment.

Vision

To create an open, extensible simulation program that unifies the language of physics, data, and computation.

Flux is meant to evolve beyond fluids:

from drone dynamics in atmospheric conditions,
to heat management in power electronics,
to data-assisted turbulence research,
and even multiphysics coupling for renewable systems.

Each new feature or solver is a step toward that larger goal, a living laboratory where equations and intelligence coexist.

The Argent Letters Philosophy

Argent Letters is more than a collection of tools — it’s a philosophy of how science and engineering should be done.
We believe that research, computation, and design should remain transparent, reproducible, and accessible.

Argent Flux is our first step in building that vision: an open, self-directed platform for simulation and discovery.
By developing it in-house and keeping it open, we ensure that our priorities stay aligned with curiosity, rigor, and integrity, not with commercial or institutional pressures.

In an era where the publish-or-perish mentality often rewards speed over substance, we aim to slow down and do good science:
to document clearly, validate thoroughly, and make every idea reusable.
Argent Letters exists to give us, and others who share that mindset, a place to build lasting, transparent work that isn’t owned, rented, or constrained by larger systems.

Flux represents flow: of energy, matter, and knowledge.

The Road Ahead

Validation of diffusion, convection, and transport models
Parallel domain decomposition (MPI/OpenMP)
Mesh and geometry handling for complex 3D cases
Data-driven turbulence and neural policy integration
Cluster scaling and GPU compute layer
Public API for external model coupling

Contributing & Collaboration

Flux is currently in active development and evolving quickly.
We welcome contributions in numerical methods, ML integration, and cluster performance.
If you’re interested in collaborating on solver development or domain-specific extensions, feel free to reach out once the public guidelines are released.

_{Argent Flux — simulation for the flow of intelligence and matter.}

README.md Unescape Escape

Argent Flux

What is Argent Flux?