Crystal Prediction Engine

2,000,000+
compositions per second

Materials discovery from atomic physics. No GPU. No training data. No DFT. Consumer hardware. 99.1% accuracy validated against experimental databases.

78.6M
candidates screened
99.1%
vs experimental databases
130+
material domains
1,070+
patent claims filed
LoNC Materials Discovery — From First Principles

Predict crystal stability before synthesizing

DFT (density functional theory) takes hours per composition. Machine learning needs training data that doesn't exist for novel materials. LoNC computes stability from atomic physics directly — formation energy, lattice parameters, electronic structure — at over 2 million compositions per second on consumer hardware.

17B×
Faster than DFT for equivalent screening
0
GPUs required. Runs on a laptop.
0
Training data needed. Physics, not ML.
2M+
compositions per second
Screen the entire chemical space for a domain in minutes, not years. Validated against experimental databases at 99.1% accuracy.
What We Screen

130+ domains. Six that change everything.

Partnered with UNLV
Nitrogen Fixation
Haber-Bosch consumes 2% of global energy and requires extreme heat and pressure to produce fertilizer. We screened 176 catalysts achieving 95% Faradaic efficiency at ambient conditions. Fertilizer without the fossil fuel cost.
176 catalysts at 95% FE, ambient conditions
Solid-State Electrolytes
Next-generation batteries. Higher energy density, no liquid electrolyte, no dendrite formation, no thermal runaway. The battery that makes EVs go 1,000 miles and charges in 5 minutes.
600,000+ candidates screened
Room-Temperature Superconductors
Zero-resistance electrical transmission at ambient conditions. Lossless power grids. Levitating transport. MRI machines without liquid helium. The most sought-after material in physics.
78.6M total candidates in pipeline
Rare-Earth-Free Magnets
Permanent magnets without neodymium or dysprosium. China controls 90% of rare earth processing. These magnets power every EV motor, wind turbine, and missile guidance system on earth.
Eliminates critical supply chain dependency
Radiation Shielding
Materials that protect humans and electronics in high-radiation environments. Space travel, fusion reactors, nuclear waste handling. Current materials are heavy and insufficient for deep-space missions.
Enabling long-duration spaceflight
Ultra-High Temp Ceramics
Materials that survive above 2000°C. Hypersonic vehicle leading edges. Rocket nozzles. Fusion reactor first walls. Reentry vehicle thermal protection. The materials that don't exist yet — but need to.
500,000+ propulsion candidates screened
PFAS Destruction Catalysts
"Forever chemicals" contaminate every water source on earth. Current destruction methods are energy-intensive and incomplete. We screen for catalysts that break the C-F bond efficiently at scale.
Targeting complete mineralization
Water Splitting
Green hydrogen requires precious metal catalysts (platinum, iridium) that cost too much to scale. We found 1,013 HER/OER catalysts without precious metals, under 50 mV overpotential.
1,013 precious-metal-free catalysts
CO₂ Capture & Conversion
Current carbon capture sorbents top out around 5 mmol/g capacity. We screened 34 sorbents at 49 mmol/g — 10x state of the art. Plus 199 electrocatalysts that convert captured CO₂ into fuel.
10× SOTA capture capacity
Performance

The numbers

2M+
Compositions/second
Consumer CPU, no GPU
99.1%
Accuracy vs experiment
Cyclical Fractal Twin verification
78.6M
Candidates screened
Across 130+ material domains

The path to Type 1

A Type 1 civilization harnesses all the energy available on its planet. That requires materials we haven't discovered yet — superconductors for lossless grids, solid-state batteries for total energy storage, fusion-grade materials for unlimited power. If this engine validates in labs, it unlocks the materials that make all of it possible. Master energy. Master materials. Everything else follows.

From atomic physics to
new materials

Lab partnerships. Government programs. Material-specific screening. Let's talk.

Get in touch