Higher-temperature superconductors would rewrite energy, computing, and medicine.
They’re hiding in chemistry today’s AI can’t search. We find them, with generative models grounded in eighteen years of physics.
Built with
Act IThe prize
Superconductors sit underneath the systems that move energy, run computation, and care for people. A better one does not improve a market. It lowers the floor under all of them at the same time.
$10T+
The energy economy, every year
Close to a tenth of all the power humanity generates is lost as heat before it ever arrives. A lossless grid reclaims it. So do lossless storage, fusion magnets, and the electrification of everything. This is the system the entire modern economy is built on top of.
$7T
The AI and quantum buildout
Every leading quantum computer is built from superconducting circuits, and the data-center buildout behind AI is tracking toward seven trillion dollars this decade. Better superconductors set the ceiling on both at once.
≈45%
Of the world's electricity runs through motors
Nearly half of all electricity drives electric motors, and the same materials carry maglev transport, MRI, and particle accelerators. Lift the temperature ceiling and every one of them steps change together.
Superconductivity is not one more materials market. It is the layer beneath the ones that already run the world. A room-temperature superconductor that can be built at scale resets the cost of all of it at once.
Global energy ≈ $10T / yr · data-center buildout ≈ $7T this decade · ~45% of electricity through motors · the figures are the systems, not the components.
Forty years of progress. Two physics families. Most of the landscape above the liquid-nitrogen line has barely been searched. MEL searches it, including the unconventional and high-Tc regimes where conventional AI cannot.
MEL hunt range · above LN₂ · novel and unconventional families
YBCO·92 K
The first cuprate to break liquid nitrogen. Workhorse of every commercial HTS magnet today.
BSCCO·110 K
Long-wire HTS used in fault-current limiters and motors. Higher Tc, harder fabrication.
Hg-1223·135 K
Ambient-pressure cuprate record. The highest Tc humanity has confirmed without extreme pressure.
LaH₁₀·250 K
The all-time record, but only stable inside a diamond-anvil cell. Not deployable.
RT·293 K
The prize. An ambient-pressure room-temperature superconductor rewrites power, computing, and medical imaging at once.
Public physics values · BCS conventional ceiling · cuprate family · ambient-pressure and high-pressure records
Act IIThe method
Three things we do that nothing else does. Together, they are why leading labs in modern condensed matter measurement have agreed to collaborate with us.
Most AI for materials is trained on simulations that break down on the materials where breakthroughs actually live. We start from a physics framework built for those systems and search inside it.
Generic ML can't see what isn't in its training set.
MEL is the only physics framework built specifically for high-temperature superconductors. Its originator developed it over eighteen years before we wrote a line of platform code. We didn't invent the science. We made it searchable.
Time you can't compress with capital.
Every prediction is measured in collaboration with research groups at UIUC and UC Berkeley, labs deeply experienced in the techniques each prediction depends on. Confirmations feed back as priors, and predictions sharpen with every cycle.
AI predictions only matter when reality validates them.
The technical comparison, DFT vs MEL, lives on /science
Read the scienceFive stages, running continuously, end to end. Click any one to step through. The loop never stops. Each closed cycle sharpens the predictions in the next.
Stage 01
Read the material's playbook
Feed the platform a candidate material. It reads the atomic structure and identifies which quantum behaviors the structure can host: superconductivity, charge order, spin order, orbital order. Think of it the way a chess engine reads a board. The moves a position permits are now known.
A live trace of the discovery pipeline, end to end: classifying candidates, generating new ones, screening, synthesizing, closing the loop with a collaborating lab. Hover any line to decode the physics behind it.
events
1,242
confirmations
47
p50 latency
1.8 s
±1 K
predicted vs measured Tc, across the known literature
Run backward over every superconductor the field has already measured, the MEL criterion reproduces each critical temperature to within a single degree. No fitting to the answer, no free parameters tuned per material.
Act IIIThe proof
In 2026, a team at Stanford and SLAC published in Physical Review Letters, one of the most selective journals in physics. They measured the effect SuperMatics is built to find: in a copper-oxide superconductor, charge order and superconductivity strengthen one another instead of competing. That cooperation is the signal our platform uses to rank candidates, and an outside group has now observed it directly.
Stanford and SLAC followed the charge order in a cuprate as it was cooled through its transition temperature, with enough resolution to see how that order held together below Tc.
Below Tc, the charge order strengthens in step with the superconducting state, and its periodicity locks onto the lattice rather than drifting. The two reinforce one another rather than compete.
This cooperative behavior is what MEL was built to describe, and the signal our platform ranks candidates on. The full formalism is on the science page.
From the scientific founder
“We have argued for years that charge order and superconductivity in cuprates reinforce one another rather than compete. The Stanford and SLAC measurement reads that same cooperative behavior out of the data, in the same family of materials. The story we built the platform on is the story the experiment tells.”
What MEL predicted, an independent group has now measured. The platform ranks candidates on the criterion this result isolates.
See the formalismEvery prediction is measured. Our collaborators run the techniques MEL was designed to be confirmed by: scanning tunneling microscopy at UIUC, synthesis at UC Berkeley, with further research discussions underway across leading correlated-electron groups.
Prof. Vidya Madhavan
University of Illinois Urbana-Champaign
University of Illinois Urbana-Champaign
STM / STS · FT-STS
Experimental collaboration · STM/STS, FT-STS
QB3 / Berkeley Nanofabrication Center
UC Berkeley · Waqas Khalid
UC Berkeley · Waqas Khalid
Synthesis · Fabrication
Research infrastructure · synthesis and fabrication
Czech Republic · HTS manufacturing
Czech Republic · HTS manufacturing
HTS synthesis · Scale-up
Industrial HTS synthesis of MEL candidates · advisory on synthetic routines and scale-up
Advanced alloys · metal powders · PVD targets
Advanced alloys · metal powders · PVD targets
PVD targets · Powders
Industrial synthesis of MEL candidates · sputtering targets and precursor powders for thin-film and bulk routes
Wilson Sonsini Goodrich & Rosati
WSGR
WSGR
Patent · Corporate
Patent strategy · IP and corporate counsel
01
Operating temperature
>120 K, ambient
Quantum computing
02
Thermal conductivity
>1500 W/m·K
AI training and inference clusters
03
Efficiency figure
>3 at 300 K
Cryogenic cooling
04
Ionic conductivity
>10⁻² S/cm
Solid-state batteries
Act IVThe work
What changed this quarter. Concrete artifacts only: a published paper, signed collaborations, a shipped build of the classifier, and the roles we’re hiring for.
Stanford and SLAC publish in Physical Review Letters (Lee et al., 2026). Resonant soft x-ray scattering on a cuprate shows CDW phase coherence growing BCS-like below Tc with near-perfect wavevector locking. The cooperative CDW-SC behavior MEL was built to predict, measured directly.
CAN Superconductors begins build of MEL-generated high-Tc candidates and advisory on synthesis routines.
Eloi Materials (EML) joins as an industrial supplier, providing PVD sputtering targets and precursor metal powders for thin-film and bulk synthesis of MEL candidates.
Eighteen years of foundational research distilled into one criterion. Back-test agrees with measured Tc within ±1 K across the public superconductor literature. arXiv:2512.03368.
Madhavan group (UIUC) · QB3 / Berkeley Nanofabrication Center for synthesis · further discussions underway with leading correlated-electron groups.
Classification, generative proposal, computational triage, and synthesis routing wired without manual handoffs. Generating candidate batches daily across the platform.
Berkeley-anchored, remote possible. Looking for builders with deep-tech instincts and the patience for hard physics.
01 · Investors
Deep-tech capital at the AI × physics interface. Brief by invitation.
Open investor brief02 · Research collaborators
Condensed matter theorists and experimentalists. If your work is at the frontier of unconventional superconductors or quantum materials, let’s talk.
research@supermatics.io03 · Builders
CEO and Solutions Architect open in Berkeley. Foundational work, ownership stakes.
See open rolesOr write to contact@supermatics.io. We read everything.