Built to discover the next generation of superconductors, fusion materials, and quantum substrates.

A simulation engine built for the physics you actually need.

The world's first computational platform that computes electronic structure from first principles without empirical corrections or material-specific parameters. Where DFT approximates, our engine computes.

Built on three decades of foundational research in geometric physics, our GPU-accelerated platform simulates electronic structure and transport from a single governing equation. The result: material property predictions in regimes where no existing tool can operate, delivered before you synthesize a single sample.

The hardest problems in materials science share one thing in common.

They all depend on strongly correlated electrons, and they all exceed the reach of existing simulation. Here is where our first-principles approach changes the game for your research.

FIRST PRINCIPLES DATA-FITTED KNOWN MATERIALS NOVEL MATERIALS Quantum Monte Carlo Publishable accuracy, but months per result. Not a design tool. DFT Optimizes what already exists. Cannot predict properties of materials never yet made. AI / Machine Learning Fast for known chemistries. Blind outside training data. Invention limited. Velar Scientific Predicts properties of materials that don't exist yet.

High-Temperature Superconductors

Our engine is designed to compute the electronic interactions that govern critical current density from first principles, replacing trial-and-error synthesis with targeted material design.

Fusion Materials

Designing plasma-facing materials that survive reactor conditions requires simulation at the electron level. Our engine is built to model the coupled electromagnetic environment that existing codes approximate.

Quantum Substrates

Topological insulator design requires computing band structures and edge states from first principles. Our engine is built to deliver the predictions that existing tools cannot.

From your target material to validated predictions in three steps.

01

Define the target

You tell us the material properties you need and the system you are studying. We scope the simulation and define the computational approach.

02

We compute

Our engine runs first-principles simulations on your candidate materials, predicting properties in regimes where existing tools cannot operate.

03

You build

You receive validated material candidates with predicted properties, ready for targeted synthesis and testing. Fewer cycles, faster discovery, less wasted spend.

Decades of aerospace engineering meets breakthrough physics.

Built by leaders in high-performance computing, numerical physics, and deep-space systems engineering.

Michael Rudolph
Founder & CEO

Co-founder of the physics research group whose theoretical work Velar’s engine is built on. Nine years at NASA’s Jet Propulsion Laboratory. MS in Astronautical Engineering, USC.

Dr. Mayank Drolia
Technical Advisor

PhD. from Heriot-Watt University, Post-doctoral research at University of Cambridge on scalable numerical methods. Expert in high performance simulation.

Arnie Benn
Science & Strategy Advisor

Expert in chemistry at the atomic scale. Co-author of published work on quantum spin topology and electron behavior.

Let's discuss what's possible.

If your team is pushing the limits of what current simulation tools can handle, reach out. We work with R&D teams under strict NDA to scope whether our engine can accelerate your materials discovery.

Download Capability Brief