# SuperMatics, in full

> Software, lab and pilot manufacturing for superconducting devices.

SuperMatics finds the materials and manufacturing processes your device needs, then makes and measures them: lab work at Berkeley, partner benches, and pilot runs with manufacturers.

Canonical website: https://supermatics.io

## Business and availability

SuperMatics is a platform infrastructure company for superconducting devices: software that finds the material and the process, lab work that makes and measures it, and pilot manufacturing with industrial partners. SuperMatics is a platform infrastructure company, not a software vendor: a program runs from the search to a sample made and measured at your device’s operating point, and on to the part with the people who build it.

Our business combines platform software licenses, paid development programs, and licensing of materials and processes we own.

Paid search campaigns running with fusion and quantum hardware teams; live with CAN Superconductors today. Our team operates the software today; customer-operated enterprise access is planned.

Customer programs address device requirements using materials and processes. The company's internal discovery program searches for its own materials. A ranked route is a proposal for testing, not a measured result.

Source: https://supermatics.io/platform and https://supermatics.io/investors

## Scientific basis and current research

Customer engineering programs can work with known superconductors. Separately, our discovery research draws on the Modulated Electron Lattice (MEL) framework, which has been in development at Hyunsung TNC since 2006 and is licensed exclusively to SuperMatics. 20+ patents issued and pending.

Five proposed materials in synthesis. Some are alloy-based candidates targeting high-temperature superconductivity. None has yet been independently established as superconducting.

An independent experiment produced a result consistent with a central MEL prediction. Stanford and SLAC performed the measurement with no involvement from SuperMatics. This does not validate the entire framework or establish a SuperMatics material as superconducting.

Superconductivity reinforces charge-density-wave phase coherence across cuprates. Lee et al., Phys. Rev. Lett., 2026. DOI 10.1103/g41t-8456. https://journals.aps.org/prl/abstract/10.1103/g41t-8456

Sources and limits: https://supermatics.io/reference#basis
Research status: https://supermatics.io/discovery

## People

- Founders: James Kim, Ph.D. (Scientific Founder · MEL Theory; CTO, Hyunsung TNC); Davis Rens (Co-founder · Chief Scientific Officer; SuperMatics · Berkeley); Charlie Moon (Co-founder · Chief Strategy Officer; SuperMatics · Hyunsung TNC).
- Officers who are not founders: Bruce Wu (Fractional Chief Legal Officer).
- Advisors: Dr. Waqas Khalid (Strategic Advisor, QB3 / Berkeley Nanofabrication Center · UC Berkeley); Dr. David Kiewlich (Advisor, BADASS Labs · Tomorrow Biotech Corporation). Advisors' affiliations are their own institutions.

James Kim, Ph.D.: James developed the MEL framework behind our discovery research and co-built the SuperMatics platform. He holds a Ph.D. in Chemistry from the Karlsruhe Institute of Technology, and developed the framework at Hyunsung TNC in Suwon, where he is chief technology officer.

Davis Rens: Davis co-founded SuperMatics, built the platform with James, and is its chief scientific officer. He took dual degrees in Chemistry and Physics at UC Berkeley in three years, and came from Rigetti Computing, where the cryogenic cabling and thermal-management software he built are still in production.

Charlie Moon: Charlie ran companies for twenty-five years before SuperMatics, fifteen of them as chief executive of KOPEX USA, the cable-protection line ABB carries. He owns the commercial side: IP, licensing, alliances, and the U.S. to Korea pipeline across SuperMatics and Hyunsung TNC.

Source: https://supermatics.io/company

## Published temperature record

These are published measurements compiled on Reference, not platform predictions or SuperMatics materials.

- 1911: Hg, 4.2 K, ambient pressure.
- 1941: NbN, 16 K, ambient pressure.
- 1954: Nb₃Sn, 18.3 K, ambient pressure.
- 1973: Nb₃Ge, 23.2 K, ambient pressure.
- 1986: La-Ba-Cu-O, 35 K, ambient pressure.
- 1987: YBCO, 92 K, ambient pressure.
- 1988: Tl-2223, 122 K, ambient pressure.
- 1993: Hg-1223, 135 K, ambient pressure.

A separate result: 2026, Hg-1223, 151 K, pressure-quenched. The pressure-quenched phase is metastable: it survives about three days at 77 K. It is not a durable conductor. DOI 10.1073/pnas.2536178123.

Source: https://supermatics.io/reference#record
Download the table: https://supermatics.io/reference/temperature-record.csv

## Testing and independent literature

The platform ranks materials and processes against engineering requirements. Our discovery research draws on the MEL framework to search for new materials; it is a distinct part of our work. Evaluate a prediction against published measurements, test it in a new experiment, and examine the independent evidence for its physical basis. Each answers a different question.

We use retrospective and prospective tests. In the retrospective direction, established superconductors are evaluated with part of the record withheld; the platform returns its result in minutes for comparison with published measurements. In the prospective direction, a prediction and its uncertainty are recorded before the instrument runs. Comparisons must account for what was measured: onset, midpoint, zero resistance, and the Meissner response are different criteria. We do not publish a precision figure that collapses those distinctions.

The studies below examine how charge order, lattice motion, and superconductivity interact. Their relevance to MEL differs: one reports an effect consistent with a central prediction; others provide related experimental observations, theoretical context, or computational precedents. Each group states what its evidence supports. None is an independent test of the full SuperMatics platform.

### Independent experimental result consistent with a central MEL prediction

Superconductivity is measured raising the phase coherence of the charge order while lowering its amplitude. MEL holds those two apart and predicts that exact split, which is what makes this the closest outside result to the framework. The group at Stanford and SLAC did not set out to test MEL, and we had no part in the work.

- Superconductivity reinforces charge-density-wave phase coherence across cuprates. Lee, H., Kuo, C.-T., Fujita, M., Kao, C.-C. & Lee, J.-S., Phys. Rev. Lett. 136, 186502 (2026). https://journals.aps.org/prl/abstract/10.1103/g41t-8456

### Charge and lattice, watched moving together in two copper oxides

In YBCO the lattice’s slow fluctuations change as the material crosses into superconductivity, and change again where the charge order sits. In LSCO the tilt of the CuO₆ octahedra tracks both the electron-phonon coupling and the charge correlations under it. MEL is a claim about an electron lattice; this is the lattice half of it.

- Understanding the superconductivity and charge density wave interaction through quasi-static lattice fluctuations. Porter, Z. et al., Proc. Natl. Acad. Sci. 121, e2412182121 (2024). https://doi.org/10.1073/pnas.2412182121
- Enhanced electron–phonon coupling due to layered-perovskites octahedral tilting in hole doped cuprate superconductor. Isha et al., J. Phys.: Condens. Matter 38, 145402 (2026). https://doi.org/10.1088/1361-648X/ae574f

### The same intertwined state, in a kagome metal

Short-range charge correlations survive across the second superconducting dome after the long-range order is gone. Short-range and surviving is the state MEL is built on, and here it is in a kagome metal rather than a copper oxide.

- Interplay between charge correlations and superconductivity across the superconducting domes of CsV₃Sb₅₋ₓSnₓ. Capa Salinas, A. N. et al., Phys. Rev. B 114, 094504 (2026). https://journals.aps.org/prb/abstract/10.1103/bcqw-7lsf

### The amplification derived independently, from another starting point

Electronic density fluctuations are shown to enhance the electron-phonon coupling and the pairing interaction that comes out of it. The amplification MEL leans on is derived here from a different starting point.

- Effective enhancement of the electron-phonon coupling driven by nonperturbative electronic density fluctuations. Moghadas, E., Reitner, M., Wehling, T., Sangiovanni, G., Ciuchi, S. & Toschi, A., Phys. Rev. Lett. 136, 126502 (2026). https://journals.aps.org/prl/abstract/10.1103/ld8v-qpf8

### A finite-q lattice instability, measured directly

A softened finite-momentum phonon is identified as the microscopic origin of charge order in CsV₃Sb₅. Finite momentum is where MEL says the physics happens, and this is a direct measurement of it.

- Soft mode origin of charge ordering in superconducting kagome CsV₃Sb₅. McGuinness, P. H. et al., Nat. Commun. 17, 4817 (2026). https://www.nature.com/articles/s41467-026-73662-4

### Modulated pairing from a neighboring theory

Kagome sublattice structure is shown to generate a modulation of superconducting pairing inside the unit cell. A different lattice, a different method, and a modulated pairing state falls out of it.

- Sublattice modulated superconductivity in the kagome Hubbard model. Schwemmer, T., Hohmann, H. et al., Phys. Rev. B 110, 024501 (2024). https://journals.aps.org/prb/abstract/10.1103/PhysRevB.110.024501

### Foundational review

The review that settled why charge, spin, nematic and pair-density-wave orders in these materials are better read as intertwined with superconductivity than as competing with it. MEL is a quantitative framework built on that reading.

- Colloquium: Theory of intertwined orders in high temperature superconductors. Fradkin, E., Kivelson, S. A. & Tranquada, J. M., Rev. Mod. Phys. 87, 457 (2015). https://doi.org/10.1103/RevModPhys.87.457

### Computational precedents, not evidence for MEL

Prediction to synthesis has been done before, and at scale: a kagome superconductor predicted and then made, and a screening workflow built on learned Eliashberg spectral functions with stability calculations and experimental follow-up behind it. This is the standard our own workflow has to beat. Neither paper is ours, and neither is evidence for the physics.

- Machine-learning-guided discovery of kagome superconductors YRu₃B₂ and LuRu₃B₂. Albu Mustaf, R. et al., Phys. Rev. Research 8, 023308 (2026). https://journals.aps.org/prresearch/abstract/10.1103/lpqj-7hyg
- Developing a complete AI-accelerated workflow for superconductor discovery. Gibson, J. B. et al., npj Comput. Mater. 12, 95 (2026). https://www.nature.com/articles/s41524-026-01964-8

Source: https://supermatics.io/reference#proof

## Notes

### The ambient-pressure superconductor record, 1911 to 2026

2026-09-14. Every material that has held the ambient-pressure transition temperature record, and what the 2026 pressure-quenched result does and does not establish.

https://supermatics.io/notes/ambient-pressure-temperature-record

Index: https://supermatics.io/notes
Feed: https://supermatics.io/feed.xml

## Published elsewhere

These are ours. A piece written by a founder is not independent coverage of this company and is not offered as any; reporting by other people is on News.

### The Last Material

On superconductors, the physical substrate of our reality, and how little we have decided to want.

Davis Rens, Substack, 2026-09-12. The first thing Davis published under his own name rather than the company's, and the argument that sits under everything here: that the material is the constraint on the things built downstream of it, and that almost nobody is funding the search for a better one. Written as himself, at length, and it says so in the first line.

https://davisrens.substack.com/p/the-last-material

## Glossary

### Superconductor

A material that carries current with no resistance, so no energy is lost as heat. Each does it only below some temperature, usually one near absolute zero.

Every use of one starts by cooling it, so the temperature a material needs sets the size, the cost and the power draw of the machine built around it.

https://supermatics.io/glossary/superconductor

### Critical temperature (Tc)

The temperature a material has to be colder than before it superconducts. It is the number the field has chased for a century, and it is quoted in kelvin.

A device team reads it as a refrigeration bill: above 77 K the coolant is liquid nitrogen, and below about 20 K it is liquid helium or a closed-cycle machine.

https://supermatics.io/glossary/critical-temperature

### Kelvin (K)

The scale that starts at absolute zero, in steps of one degree Celsius. Room temperature is about 293 K, liquid nitrogen boils at 77 K, liquid helium at 4.2 K.

Every number on a conductor datasheet is quoted at a temperature, so two figures given without their kelvin are not comparable.

https://supermatics.io/glossary/kelvin

### Ambient pressure

Ordinary atmospheric pressure, the pressure everything a person can build operates at. A material that only superconducts under enormous pressure is a real result and not yet a usable one.

It is the line between a published result and a buildable one, and it is why the pressure a measurement was taken at is quoted beside the temperature.

https://supermatics.io/glossary/ambient-pressure

### Meissner effect

A superconductor pushes magnetic fields out of itself, so a magnet floats above one. It is the definitive test: an ordinary conductor can fake zero resistance and cannot fake this.

It is what separates a superconductor from a very good conductor, so a measurement without it is not yet a claim of superconductivity.

https://supermatics.io/glossary/meissner-effect

### Cooper pair

Two electrons bound together so they move through the material as one unit. Superconductivity is current carried by these pairs, which travel without scattering their energy away as heat.

The pairing is the thing a theory of superconductivity has to explain, and it is where the copper oxides still resist explanation.

https://supermatics.io/glossary/cooper-pair

### BCS theory

The 1957 theory of every superconductor then known: lattice vibrations bind electrons into pairs. It won a Nobel Prize, and does not explain the copper oxides holding the ambient-pressure record.

It implies a ceiling on transition temperature that the copper oxides went past, so the record holders are the materials the settled theory does not cover.

https://supermatics.io/glossary/bcs-theory

### Unconventional superconductor

A superconductor whose pairing BCS theory does not explain. The cuprates are the famous family, and the field’s central fact is here: highest temperatures come with the least settled physics.

The highest transition temperatures sit in this group, so the record holders and the least settled physics are the same materials.

https://supermatics.io/glossary/unconventional-superconductor

### High-temperature superconductor (HTS)

The families that superconduct above roughly 30 K, and among them the cuprates that run on liquid nitrogen rather than helium. The refrigerant changes the economics of everything built on them.

Liquid nitrogen costs a fraction of liquid helium and needs a fraction of the plant, so crossing 77 K changes what a project can afford to build.

https://supermatics.io/glossary/high-temperature-superconductor

### Type II superconductor

A superconductor that admits a strong magnetic field as threads of flux instead of failing outright. Every practical conductor is Type II, since a magnet must survive its own field.

A magnet has to survive the field it makes, so every conductor wound into one is Type II.

https://supermatics.io/glossary/type-ii-superconductor

### Critical current density (Jc)

How much current a superconductor carries per unit cross-section before it stops superconducting. A high transition temperature with a low critical current is not a conductor anybody can build with.

It decides how much conductor a given current needs, and with it the mass, the cost and the size of the finished magnet.

https://supermatics.io/glossary/critical-current-density

### Critical field

The magnetic field a superconductor tolerates before it gives up. Most things people want superconductors for are magnets, so the material has to survive its own field.

A magnet's own field is the load its conductor has to tolerate, so critical field is read at the operating temperature rather than on its own.

https://supermatics.io/glossary/critical-field

### Cost per kiloamp-meter ($/kA·m)

The price of moving a thousand amps across one meter of conductor. It decides whether a magnet gets built, and it only means something at a stated temperature and field.

It is the number a program is approved or refused on, and it compares between materials only when the temperature and the field are stated with it.

https://supermatics.io/glossary/cost-per-kiloamp-meter

### Cryocooler

A closed-cycle refrigerator that keeps a superconductor cold without liquid refills. The colder the conductor, the larger the machine and its power draw. A warmer superconductor means a smaller one.

It is usually the largest and most power-hungry part of a superconducting system, so the temperature the conductor needs sets the machine's footprint.

https://supermatics.io/glossary/cryocooler

### Cuprate

A family of copper-oxide ceramics that includes every superconductor that runs on liquid nitrogen. They were discovered in 1986 and they still hold the ambient-pressure record.

They are brittle ceramics, so turning one into something a magnet can be wound from is a manufacturing problem as much as a materials one.

https://supermatics.io/glossary/cuprate

### REBCO

A cuprate, rare-earth barium copper oxide, and the conductor the highest-field magnets are now wound from. It comes as coated tape rather than wire, which is part of its price.

It is the conductor most high-field magnet programs are built around today, and its price per kiloamp-meter is set by the tape process rather than by the compound.

https://supermatics.io/glossary/rebco

### Hydride superconductor

A hydrogen-rich compound superconducting within a few tens of kelvin of freezing, but only inside a diamond anvil cell at 150 to 200 gigapascals. Nothing can be built at that pressure.

The pressures involved are held inside a cell smaller than a grain of rice, so these results inform theory rather than devices.

https://supermatics.io/glossary/hydride-superconductor

### Metastable

Holding a state it will eventually leave. A metastable superconducting phase can be real, measured and published, and still unusable, because it does not last or cannot be made large.

A metastable phase can be measured, published and entirely real, and still not last long enough or scale large enough to be wound into anything.

https://supermatics.io/glossary/metastable

### Doping

Adding or removing charge carriers, usually by substituting a few percent of one element for another. An undoped cuprate does not superconduct at all; doped, it holds the record.

The same parent compound can be an insulator or a record holder depending on it, so a material is not specified until its doping is.

https://supermatics.io/glossary/doping

### Synthesis route

The recipe for making a material: what goes in, at what temperature, in what atmosphere, how long. A predicted compound with no route is a suggestion rather than a candidate.

A candidate with no route cannot be made, so a search that returns compositions alone returns work rather than answers.

https://supermatics.io/glossary/synthesis-route

### Coated conductor

A superconductor grown as a film on metal tape rather than drawn as wire. It makes brittle ceramic into something that bends, and the process is most of the price.

The superconducting film is a thin fraction of the tape. The substrate, the buffer layers and the metal around them are most of the cost and all of the mechanical behavior.

https://supermatics.io/glossary/coated-conductor

### Flux pinning

The holding of magnetic flux lines in place by defects engineered into a conductor, since flux that moves dissipates energy and spoils the zero resistance. Processing matters as much as composition.

Two samples of one composition can carry very different currents because of it, so a formula on its own does not predict what a conductor will do.

https://supermatics.io/glossary/flux-pinning

### Quench (in a magnet)

The moment a magnet stops superconducting and its stored energy becomes heat at once. It is unrelated to pressure-quenching a sample, and designers spend most of a design surviving it.

Surviving it drives much of a magnet's design, and the stored energy released is why it is treated as a safety case rather than an inconvenience.

https://supermatics.io/glossary/quench

### Modulated Electron Lattice (MEL)

The physics under this platform, in development at Hyunsung TNC since 2006, licensed exclusively to SuperMatics. The effect underneath it was measured independently.

It is the framework this platform runs on. What it rests on, and the limits of that, are set out with the published work on the reference page.

https://supermatics.io/glossary/modulated-electron-lattice

### Retrodiction

Testing a method on something known by hiding the answer, running it and comparing. It shows agreement with the published record, and we never call that predicting something unmeasured.

It is how a method is tested without waiting for new measurements, and it says nothing about anything that has not been measured.

https://supermatics.io/glossary/retrodiction

Index: https://supermatics.io/glossary

## Superconductors in service

These are published values for materials in service today, compiled from the literature. None of them is a SuperMatics material or a platform prediction.

### NbTi

A ductile niobium-titanium alloy, and the only superconductor that can be drawn into wire on machinery a copper mill would recognize.

- Common name: niobium-titanium
- Transition temperature: 9.2 K
- Made as: wire
- In service: MRI magnets; the LHC main dipoles; ITER’s poloidal field coils
- Source: Textbook value; CERN LHC Design Report vol. 1; ITER magnet system documentation

Every other conductor on this list is chosen in spite of how it is made. NbTi is chosen because of it. The alloy takes cold work. A billet of niobium-titanium rods packed in a copper matrix can be extruded and drawn in one continuous process, down to filaments a few microns across. No reaction furnace in the middle. No strain limit the winder has to design the magnet around. A coil shop can wind it, twist it, joint it and repair it. The cheapest superconducting wire in the world is also the one inside most of the MRI scanners installed today. A hospital magnet is a procurement decision rather than a physics program.

Where it stops: It runs out of field long before it runs out of anything else. Past the operating point a modern high-field magnet asks for there is no margin left in the alloy, and a machine that needs more does not get there by winding more turns. It changes material, which is the entire reason the industry tolerates the conductor listed next to it.

https://supermatics.io/materials/nbti

### Nb₃Sn

A brittle intermetallic with roughly twice the transition temperature of NbTi, and a manufacturing process built entirely around the fact that it cannot be bent.

- Common name: niobium-tin
- Transition temperature: 18.3 K
- Made as: wire
- In service: ITER’s toroidal field coils; the HL-LHC 11 T dipoles; high-field NMR magnets
- Source: Textbook value; ITER magnet system documentation; CERN HL-LHC technical design report

Nb3Sn is where a magnet program goes when NbTi runs out of field, and what it costs is the freedom to treat the conductor as wire. The compound only forms after a long reaction heat treatment, and once formed it is ceramic-brittle, so the coil is wound from a ductile precursor and then the entire wound coil goes into a furnace. Everything downstream inherits that. The insulation has to survive the reaction. The support structure has to be in place before it. And the finished winding is strain-sensitive enough that the Lorentz force the magnet generates on itself degrades the conductor carrying it. A Nb3Sn magnet is a mechanical design problem with a superconductor inside it.

Where it stops: The transition temperature is generous next to NbTi and low next to a brief written at twenty kelvin, which is where the fusion and high-field magnet work now sits. It is the conductor most often eliminated by the temperature line rather than by the field line.

https://supermatics.io/materials/nb3sn

### REBCO

A layer of rare-earth barium copper oxide a few microns thick, grown on a metal tape, and the conductor that made a compact fusion magnet credible.

- Common name: rare-earth barium copper oxide
- Transition temperature: 92 K
- Made as: tape
- In service: High-field fusion magnets (the 20 T SPARC toroidal-field model coil, 2021); superconducting power cables
- Source: Textbook value for YBCO; Commonwealth Fusion Systems and MIT PSFC, 20 T large-bore demonstration, 2021

REBCO is the only conductor in service that still carries useful current at the field and temperature a compact tokamak is designed around, and the twenty-tesla model coil built in 2021 is the reason anyone believes that point is reachable. It is also barely a wire. The superconductor is a coating on a hundred times more tape, so the current flows in a plane. The critical current depends on which way the field crosses that plane. The tape cannot be transposed the way a round strand can. And a normal zone spreads along it slowly, so noticing a quench before it does damage becomes a design problem in its own right.

Where it stops: REBCO almost never fails on physics. It fails on price per kiloamp-meter and on how many kilometers a supplier can deliver this year, which is a different kind of no and the one that stops the most programs.

https://supermatics.io/materials/rebco

### MgB₂

Magnesium diboride: two cheap, light, abundant elements that turned out in 2001 to superconduct far warmer than any other simple compound.

- Common name: magnesium diboride
- Transition temperature: 39 K
- Made as: wire
- In service: Conduction-cooled MRI without a liquid-helium bath
- Source: Nagamatsu et al., Nature 410, 63 (2001); ASG Superconductors / Paramed MROpen MRI

MgB2 is the conductor that takes liquid helium out of the room. Its transition temperature is high enough to run a magnet on a closed-cycle cryocooler with no bath, which means no cryogen deliveries, no recovery plant, and no quench that vents into a hospital corridor. That is not a performance argument, it is an operating-cost and siting argument, and it is why the first helium-free MRI scanners are wound from it. The material is also magnesium and boron, with no rare earth and no silver in it, so the raw ingredients are not the reason it is expensive.

Where it stops: The temperature headroom does not convert into field headroom. Run it warm and the current it will carry under field falls away, so the operating point that makes the cryogenics easy is the one where the magnet is weakest. Every design using MgB2 is a negotiation between those two facts.

https://supermatics.io/materials/mgb2

### Bi-2212

The one high-temperature superconductor that can be made as a round, isotropic wire, and the reason it exists inside the strongest research magnets in the world.

- Common name: bismuth strontium calcium copper oxide
- Transition temperature: 85 K
- Made as: wire
- In service: Round-wire inserts for magnets above 25 T
- Source: NHMFL high-field insert coil program; Larbalestier et al., Nature Materials 13, 375 (2014)

Every other high-temperature conductor is a tape, and a tape cannot be cabled and transposed the way a round strand can. Bi-2212 can. That single geometric fact is what puts it inside the highest-field magnets built. An insert coil wound from round wire behaves, mechanically and electrically, like the Nb3Sn and NbTi coils around it. The designer does not have to solve the tape problem a second time inside the same magnet.

Where it stops: The wire only reaches its published numbers after a reaction under high gas pressure, in a furnace built for the purpose. That is a capital item rather than a recipe, and the number of facilities on earth that can run it is the real ceiling on how much Bi-2212 magnet exists.

https://supermatics.io/materials/bi-2212

### Nb

Elemental niobium, which has the highest transition temperature of any element at ambient pressure, and the material most superconducting circuits are wired in.

- Common name: niobium
- Transition temperature: 9.3 K
- Made as: film
- In service: Transmon circuit wiring and resonators; accelerator RF cavities
- Source: Textbook value; the transmon literature from Koch et al. (2007) onward

Niobium is what a superconducting circuit is made of everywhere the circuit is not the junction. It patterns cleanly. It survives the process temperatures a fab wants to use. Its surface can be prepared to the point where a resonator holds energy for a very long time. And it is a single element, so there is no stoichiometry to hold constant across a wafer. The same properties put it inside accelerator cavities, where the figure of merit is how little of the drive power the wall absorbs.

Where it stops: Niobium still needs a helium bath or a dilution refrigerator, which is acceptable for the machines that use it and disqualifying for everything else. Niobium is not a candidate for anything that has to leave a laboratory.

https://supermatics.io/materials/niobium

### Al

Aluminum superconducts at close to the lowest useful temperature on this list, and nearly every superconducting qubit in the world is built from it anyway.

- Common name: aluminum
- Transition temperature: 1.2 K
- Made as: film
- In service: The Josephson junctions in transmon qubits (Al/AlOx/Al)
- Source: Textbook value; the transmon literature from Koch et al. (2007) onward

The reason is oxide. Aluminum grows a thin native oxide that stops at a reproducible thickness, and a Josephson junction is two superconductors separated by a barrier whose thickness sets the behavior of the whole device. Aluminum, its own oxide, and aluminum again gives a junction one fab can make twice. Nothing else on this list offers that, and a processor is hundreds of junctions that all have to agree with each other, so reproducibility beats transition temperature by a wide margin. The rest of the chip can be niobium; the junction is aluminum.

Where it stops: Its transition temperature is not a temperature, it is a budget item. Everything built on aluminum lives at the bottom of a dilution refrigerator, where the available cooling power is measured in microwatts. That ceiling shapes the architecture above it. How many control lines can enter the cold stage. How much can be dissipated reading a qubit out. And how large a processor gets before the refrigerator, rather than the physics, is the limit.

https://supermatics.io/materials/aluminum

### TiN

A hard, chemically stable nitride whose transition temperature can be tuned during deposition, used where a resonator has to be small and quiet.

- Common name: titanium nitride
- Transition temperature: 4.5 K
- Made as: film
- In service: High-quality-factor superconducting resonators; kinetic-inductance detectors
- Source: Vissers et al., Applied Physics Letters 97, 232509 (2010)

TiN is chosen for two properties that have nothing to do with carrying current. It has a large kinetic inductance, so a resonator made from it can be physically small and still resonate low, which is what a kinetic-inductance detector array needs. And its loss at the single-photon powers a qubit readout works at is low enough to compete with the best niobium surfaces. Its transition temperature moves with nitrogen content and strain, which is unusual on this list: the film can be tuned toward the application rather than accepted as it comes.

Where it stops: Tunability is also the exposure. A property set by deposition conditions is a property that belongs to one chamber on one day, and TiN is the standing example of a film whose published numbers are hard to reproduce somewhere else. That is a process-window and interface problem rather than a physics one, which is the class that stops programs quietly.

https://supermatics.io/materials/titanium-nitride

### NbN

A thin film with a high transition temperature for its class and a very short recovery time, which is what a single-photon detector is actually built around.

- Common name: niobium nitride
- Transition temperature: 16 K
- Made as: film
- In service: Superconducting nanowire single-photon detectors
- Source: Gol’tsman et al., Applied Physics Letters 79, 705 (2001)

A superconducting nanowire single-photon detector works by sitting just under its critical current, so that one photon carries enough energy to push a segment normal and produce a voltage pulse. What matters is not how cold the wire has to be but how quickly it recovers and re-arms. In NbN the excited electrons hand their energy to the lattice quickly, so the detector resets in nanoseconds and can count at high rates. Its transition temperature also sits comfortably above what a closed-cycle cryocooler reaches, which the designer spends as margin rather than as performance.

Where it stops: The detector is a wire a few nanometers thick and a fraction of a micron wide, and it performs like the worst constriction anywhere along its length. Making one is a film problem. Making a thousand on a chip that all work is a uniformity problem, and that, not sensitivity, is where the array count stops.

https://supermatics.io/materials/nbn

### NbTiN

Niobium nitride with titanium substituted into it, traded for a lower and better-controlled kinetic inductance and a cleaner microwave surface.

- Common name: niobium titanium nitride
- Transition temperature: 15 K
- Made as: film
- In service: Single-photon detectors and high-kinetic-inductance resonators
- Source: The SNSPD literature from Dorenbos et al. (2008) onward

NbTiN exists because the kinetic inductance of NbN is high and hard to hold constant, and in a detector array or a high-Q resonator the inductance is the quantity being engineered. Titanium lowers it and makes it repeatable across a wafer without giving up much transition temperature, and the resulting film loses less at microwave frequencies. It is the choice when the device is many elements that have to match each other, rather than one element that has to be fast.

Where it stops: It is still a nitride film whose properties are set in a deposition chamber, so it inherits the reproducibility problem of NbN in a milder form. And its transition temperature keeps it a cryocooler material, with nothing about it that changes where the cold has to come from.

https://supermatics.io/materials/nbtin

Index: https://supermatics.io/materials

## Where superconductors are used

### MRI

In service today. Most scanners are superconducting magnets. More than 30,000 are installed.

Source: Niobium-titanium and affordable MRI · Physics in Perspective. https://link.springer.com/article/10.1007/s00016-015-0172-x

An MRI image is built by holding a very uniform, very stable magnetic field across a person for minutes at a time. A resistive magnet strong enough would dissipate megawatts and drift with its own heating. A superconducting coil, once charged, is a closed loop with no resistance in it: the current keeps running, the field holds, and the electricity bill is the cooling rather than the field. That is the whole argument, and it is why MRI is the one place a superconductor is already a commodity rather than a research result.

What is stopping it: Helium. The magnet is cheap next to the cryogenics and the siting rules that a bath of liquid helium brings with it, and a quench vents it. Every serious attempt to make MRI cheaper or more portable is aimed at the cryogen rather than at the field. So the conductor to watch here is not the one with the highest transition temperature. It is the one that runs without a bath.

https://supermatics.io/applications/mri

### Particle therapy

In service today. A superconducting gantry brought a heavy-ion treatment room from 600 tonnes down to about 300.

Source: A compact superconducting rotating gantry · NIRS. https://pmc.ncbi.nlm.nih.gov/articles/PMC3941506/

Treating a tumor with carbon ions means bending a beam far stiffer than a proton beam, and rotating the whole bending system around the patient so the dose can be delivered from any angle. With resistive magnets that rotating structure is a building. Superconducting magnets bend harder in less space, and the gantry becomes a machine that fits into a hospital rather than a hospital built around a machine. The mass figure is the argument: it is not a performance number, it is a construction-budget number.

What is stopping it: A gantry magnet is ramped and rotated, not parked. That means changing fields and mechanical cycling, inside a cryostat that has to keep working while it turns. It is a much harder problem than a magnet charged once and left alone. It is also the reason a gantry of this kind is still rare while the superconducting magnet next door in radiology is not.

https://supermatics.io/applications/particle-therapy

### NMR spectrometers

In service today. The magnet that resolves a protein structure runs at a few kelvin.

Source: The first 1.2 GHz protein NMR data · Bruker. https://www.bruker.com/en/news-and-events/news/2019/bruker-announces-worlds-first-1-2-ghz-high-resolution-protein-nmr-data.html

NMR resolution scales with field, and structural biology has spent fifty years buying resolution by buying tesla. The highest-field spectrometers pass the point where niobium-titanium has anything left, so the inner coils are niobium-tin and, at the very top, high-temperature conductor. This is the application that pulls a new conductor into production before anyone else can afford it. That makes it a useful leading indicator for the magnet industry as a whole.

What is stopping it: Homogeneity and drift, held to an extraordinary tolerance in a magnet that is never switched off. A conductor that satisfies that needs joints good enough to persist for years, which is a far stricter requirement than carrying the current in the first place, and it is one of the reasons high-temperature conductors have been slow to arrive here.

https://supermatics.io/applications/nmr-spectrometers

### Biomagnetic sensing

In service today. SQUIDs read the magnetic field of brain activity through the skull.

Source: Superconducting magnetometers for brain investigations · PMC. https://pmc.ncbi.nlm.nih.gov/articles/PMC12349212/

The magnetic field produced by activity in the brain is a minute fraction of the earth’s own, and measuring it through the skull was impossible until superconducting quantum interference devices existed. A SQUID is a superconducting loop interrupted by Josephson junctions, and its output swings through a full cycle for one quantum of magnetic flux threading the loop. Its sensitivity therefore comes from a physical constant rather than from an amplifier. Nothing built on conventional electronics has caught up.

What is stopping it: Everything around the sensor. A magnetically shielded room and a helium dewar held over a patient’s head are the cost, not the SQUID, and that is what has kept the technique inside research hospitals. The competition is not a better superconductor: it is a class of room-temperature optical magnetometer that needs no cryogen at all.

https://supermatics.io/applications/biomagnetic-sensing

### Transmission cable

In service today. In the ground under Shanghai, Essen and Long Island. Rare because of the cold.

Source: A kilometer of superconducting cable under Essen · KIT. https://www.kit.edu/kit/english/pi_2014_15058.php

A superconducting cable carries several times the power of a copper circuit through a fraction of the trench, at a voltage low enough to skip a transformation stage. In a dense city the expensive thing is the right of way rather than the conductor, and that arithmetic can work. Every installed example is under a city. The losses that vanish in the conductor reappear as a refrigeration bill, so the case is made by weighing that bill against the cost of digging.

What is stopping it: The cryostat is the product. A cable is a vacuum-jacketed pipe of liquid nitrogen with a superconductor inside it, buried, for kilometers, with refrigeration plant at intervals and no tolerance anywhere along its length for a warm spot. Utilities buy assets with forty-year lives and no moving parts. This one has a plant attached to it.

https://supermatics.io/applications/transmission-cable

### Fault-current limiters

In service today. Goes normal in a millisecond and swallows a short circuit.

Source: Fault-current limiters against a real short circuit · Energies. https://www.mdpi.com/1996-1073/18/19/5302

This is the one application that uses a superconductor’s failure as the feature. Below the design current the device is invisible: no resistance, no loss, no voltage drop. Above it the conductor goes normal on its own and becomes a resistance in the path of the fault, limiting it before a breaker has finished deciding what to do. Nothing else in a grid does that passively, and it lets a utility add generation to a network whose existing switchgear would otherwise have to be replaced.

What is stopping it: Recovery. The element has to go normal, absorb the fault energy without damaging itself, and cool back down fast enough to be ready for the next one. That thermal cycle is what sets the amount of conductor in the device, and therefore the price.

https://supermatics.io/applications/fault-current-limiters

### Wind generators

Built and demonstrated. A 3.6 MW superconducting rotor ran on a Danish turbine for 650 hours.

Source: A superconducting rotor on a 3.6 MW turbine · EcoSwing. https://www.iwes.fraunhofer.de/en/research-projects/finished-projects-2019/ecoswing.html

Offshore turbine power is limited by what can be hung on top of a tower, and a conventional direct-drive generator for a large machine is a several-hundred-tonne ring of copper and permanent magnet. A superconducting rotor makes the same field in a fraction of the mass, and the saving propagates down through the tower, the foundation and the installation vessel. The demonstration on a real turbine is the evidence that matters here: it moved the idea out of the category of things that work on a bench.

What is stopping it: A cryogenic rotating machine, on a tower, in the North Sea, serviced by boat. The field winding has to survive rotation, vibration and thermal cycling for twenty years with a cold head attached, and an industry whose economics depend on availability has very little appetite for a new failure mode at the top of an offshore tower.

https://supermatics.io/applications/wind-generators

### Magnetic energy storage

Built and demonstrated. Current parked in a loop and handed back in milliseconds, for power quality.

A charged superconducting loop holds energy as current and hands it back as fast as the power electronics can switch, with no chemistry in the path and effectively no cycle limit. For power quality, where the job is to ride through a voltage sag lasting a few cycles, that response beats anything electrochemical. The energy density is poor, so the application is short and deep rather than long and shallow, and it was always going to be a niche defined by response time.

What is stopping it: Energy per dollar. A coil storing a useful number of megajoules is a large magnet with large forces on it and a refrigerator beside it, and it has spent the last two decades competing against batteries whose cost fell by an order of magnitude. This one has been outrun by its competition rather than stopped by its physics.

https://supermatics.io/applications/magnetic-energy-storage

### Quantum computers

Built and demonstrated. Most of the largest processors are superconducting circuits, held near absolute zero.

Source: Who is leading superconducting quantum computing · EPJ Quantum Technology. https://link.springer.com/article/10.1140/epjqt/s40507-025-00405-7

A superconducting qubit is a resonant circuit made non-linear by a Josephson junction, so its energy levels are unevenly spaced and two of them can be addressed as a bit. It is built by lithography, and that is why this approach scaled before the others: the fabrication is a variant of something the semiconductor industry already knows how to do. The circuit has to be superconducting because resistance is decoherence, and it has to be very cold because the spacing between those levels corresponds to a temperature well under one kelvin.

What is stopping it: Nothing in this field is limited by a transition temperature. It is limited by loss at interfaces: the oxide on a surface, the substrate under a film, the residue at an etched edge. Coherence has improved by orders of magnitude over twenty years almost entirely through materials and process work on those interfaces, and that is the constraint class, not the conductor.

https://supermatics.io/applications/quantum-computers

### Photon counting

In service today. Superconducting nanowires count single photons for quantum links and lidar.

Source: Nanowire detectors at 98 percent efficiency · NIST. https://www.nist.gov/publications/superconducting-nanowire-single-photon-detectors-98-system-detection-efficiency-1550-nm

A nanowire held just below its critical current is a photon detector with almost no dark count, timing precision in the tens of picoseconds and efficiency close to unity at telecom wavelengths, and no semiconductor detector achieves all three at the same time. Quantum key distribution, deep-space optical links and single-photon lidar all run on them, and in each case the reach of the whole system is set by how few photons the receiver can work with.

What is stopping it: Scale. One detector is routine and an array of thousands is not, because the nanowire’s yield depends on uniformity along its entire length and every channel has to be read out of a cryostat through a wire that is also a heat leak. Array count, not sensitivity, is the open problem.

https://supermatics.io/applications/photon-counting

### Superconducting logic

Waiting on a material. Switching at a fraction of the energy of silicon. Demonstrated for decades, never volume-made.

Source: Superconducting logic and what it costs to switch · PMC. https://pmc.ncbi.nlm.nih.gov/articles/PMC6642163/

A single-flux-quantum circuit represents a bit as one quantum of magnetic flux moving between Josephson junctions. Switching costs a small fraction of what a silicon gate costs and the circuits clock in the tens of gigahertz. Working processors have been built and measured. The physics has never once been the obstacle here, which is what makes this the most frustrating entry on the list.

What is stopping it: Memory and manufacturing. There is no dense, fast, cryogenic memory in any material to put beside the logic, so the architecture spends its energy advantage moving data out to warm memory, and no foundry runs Josephson junctions at the volumes and yields a product would need. The thing that is missing is a material and an ecosystem at once.

https://supermatics.io/applications/superconducting-logic

### Telescope detectors

In service today. Sensors that measure one photon by the heat it leaves behind.

Source: The SPT-3G focal plane, 16,000 detectors · arXiv. https://arxiv.org/abs/1902.09640

At millimeter and submillimeter wavelengths the useful detector is a thermometer: absorb the photon, measure the temperature rise. A superconductor sitting on its transition is the most sensitive thermometer available, because its resistance changes steeply there for a very small change in temperature. That is a transition-edge sensor. A kinetic-inductance detector does the same job by watching a resonator shift in frequency instead, which has the practical advantage that thousands of them can share one readout line by each being given a different frequency.

What is stopping it: Focal planes are limited by how many channels can be read out of a cold stage, not by how sensitive one pixel is. Every wire into a cryostat carries heat in, so the multiplexing factor is the figure of merit, and improving it is a materials and readout problem rather than a detector problem.

https://supermatics.io/applications/telescope-detectors

### Maglev

Built and demonstrated. Japan’s superconducting maglev reached 603 km/h on a test track.

Source: The superconducting maglev · JR Central. https://global.jr-central.co.jp/en/company/_pdf/superconducting_maglev.pdf

Electrodynamic suspension levitates a vehicle by inducing currents in coils in the guideway as it passes, and that requires a very strong magnet on board. What it gives back is a levitation gap measured in centimeters rather than millimeters, which makes the system tolerant of track settlement and of earthquakes. That gap, not the top speed, is the engineering reason for the superconductor. The speed record is the demonstration.

What is stopping it: Civil engineering rather than conductor. The guideway is the cost, and it is an entirely new right of way for a vehicle that cannot run on existing track. The onboard cryogenics have now been solved twice over, first with niobium-titanium and later with high-temperature conductor, and neither solution made the route any cheaper to build.

https://supermatics.io/applications/maglev

### Electric aircraft

Waiting on a material. A megawatt motor light enough to fly needs a cryostat that survives a wing.

Source: The high-efficiency megawatt motor · NASA. https://www.nasa.gov/eap-technology/electric-machines/hemm/

Electrifying a regional aircraft needs a propulsion motor in the megawatt class at a specific power several times what conventional machines deliver, because every kilogram of propulsion is a kilogram not carried as payload or battery. Superconducting windings make the required field in far less mass, and every credible study of the configuration includes them. No aircraft flies on one.

What is stopping it: The cold. A flight-weight cryogenic system that survives the vibration, the altitude cycling and the certification regime of an airframe does not exist, and the mass saved in the motor has to be handed straight back to the cryostat. Until a conductor works warm enough to change that arithmetic, this stays a design study.

https://supermatics.io/applications/electric-aircraft

### Ship propulsion

Built and demonstrated. A 36.5 MW superconducting motor was built and tested for naval use.

Source: A 36.5 MW propulsion motor at full power · AMSC. https://ir.amsc.com/news-releases/news-release-details/amsc-and-northrop-grumman-announce-successful-load-testing-365/

A propulsion motor sits low and far aft where a hull has least room, and its size drives the arrangement of everything around it. A superconducting motor of the same power occupies a fraction of the volume and mass, which is a naval architecture argument rather than an efficiency one: it buys deck space and flexibility in where the machinery goes. One was built and load-tested at full power, so feasibility is settled.

What is stopping it: A ship is a poor place for liquid cryogen and a good place for shock loading. A fleet requires the cooling plant to be as reliable as the diesel it replaces, over a service life measured in decades and far from a laboratory, and no program has been willing to be first.

https://supermatics.io/applications/ship-propulsion

### Particle accelerators

In service today. 1,232 superconducting dipoles bend the beam around 27 km at CERN.

Source: The LHC’s superconducting dipoles · CERN. https://home.cern/science/engineering/pulling-together-superconducting-electromagnets/

The energy of a circular collider is set by the product of its radius and its bending field. Radius is a civil-engineering budget; field is a conductor. The LHC reaches the energy it does because niobium-titanium, held just under two kelvin, makes the field it does over twenty-seven kilometers. Every proposal for a higher-energy machine is in practice an argument about one thing: which conductor can make more field affordably, over a longer ring. The cavities that accelerate the beam are superconducting for a separate reason: almost none of the drive power is absorbed by the cavity wall.

What is stopping it: Cost per unit of bending, across tens of kilometers. The next machine’s conductor decision is between far more niobium-tin and high-temperature tape, and it is being argued on price and on how many thousand kilometers a supply chain can produce in a decade, not on whether either one works.

https://supermatics.io/applications/particle-accelerators

### Fusion magnets

Built and demonstrated. Twenty tesla, and the reason a compact tokamak is credible at all.

Source: A 20 tesla magnet for a compact tokamak · MIT. https://news.mit.edu/2021/MIT-CFS-major-advance-toward-fusion-energy-0908

Fusion power in a tokamak rises very steeply with magnetic field, far faster than it rises with the size of the machine, so doubling the field is worth much more than building bigger. That relationship is why a twenty-tesla high-temperature coil reset the schedule of an entire industry when it was demonstrated: it moved the design point from a machine the size of ITER to one that can be built in a few years. It is the clearest example on this site of a materials result changing what an industry believes is possible.

What is stopping it: Tape, in quantity, at a price. A compact tokamak’s magnets consume thousands of kilometers of high-temperature conductor, and world production is the binding constraint on how many machines get built. Behind that sits quench detection: a magnet of this kind stores enormous energy and spreads a normal zone slowly, so noticing that one has started is a research program in its own right.

https://supermatics.io/applications/fusion-magnets

### Magnetic separation

In service today. High-gradient magnets pull iron out of clay at industrial rates.

Source: Superconducting separation of kaolin · Clay Minerals. https://www.cambridge.org/core/journals/clay-minerals/article/abs/improvement-of-brightness-of-kaolin-by-superconducting-magnetic-separation-and-characterization-of-the-impurities/2F32FBA4172FB1C8647638F49CD99FB6

High-gradient magnetic separation pulls weakly magnetic particles out of a slurry by passing it through a steel matrix inside a strong field. The field has to be held continuously on a production line, and with a resistive magnet that is a permanent and substantial electricity bill. A superconducting separator holds it for the cost of cooling instead, and the economics of brightening clay at industrial rates depend on exactly that difference.

What is stopping it: Nothing about the material. It works, it is installed, and the constraint is capital cost and the availability of people who can keep a cryogenic system running inside a minerals plant. That is a different kind of problem from every other row on this list, and worth noticing: not every stalled application is waiting on physics.

https://supermatics.io/applications/magnetic-separation

### Induction heating

In service today. Heats a metal billet far more efficiently than a copper coil.

Source: Induction heaters with high-Tc magnets · SN Applied Sciences. https://link.springer.com/article/10.1007/s42452-018-0073-0

Heating an aluminum billet before extrusion is normally done by inducing current in it with an alternating copper coil, and a large share of the electricity goes into heating the coil rather than the billet. The superconducting version inverts the problem: hold a strong steady field with a superconducting magnet and rotate the billet inside it. The billet sees a changing field and heats. The magnet dissipates nothing. The efficiency difference is large enough to pay for the cryogenics in a plant running continuously.

What is stopping it: Mechanical, mostly. Spinning a heavy metal billet inside a magnet bore at production rates is a machine-design problem, and the payback only works on a duty cycle that a high-volume extrusion line has and a job shop does not.

https://supermatics.io/applications/induction-heating

Index: https://supermatics.io/applications

## Frequently asked questions

### What is SuperMatics?

SuperMatics is a platform infrastructure company for superconducting devices: software that finds the material and the process, lab work that makes and measures it, and pilot manufacturing with industrial partners. A team brings its requirements: current under field, an interface, a process window, or cost per meter. The platform ranks proposed materials and manufacturing processes against that brief; the selected route is then made and measured at Berkeley and on partner benches, and carried to pilot scale. These programs can work with known superconductors. Our own search for new materials draws on the Modulated Electron Lattice (MEL) framework, licensed exclusively from Hyunsung TNC.

### Is SuperMatics a software company?

No. SuperMatics is a platform infrastructure company, not a software vendor: a program runs from the search to a sample made and measured at your device’s operating point, and on to the part with the people who build it. The software is where a program starts. The lab work runs at Berkeley, where synthesis is moving in-house, and on partner benches at UIUC and Georgia Tech. CAN Superconductors, an industrial manufacturer, carries a route to pilot scale and advises on making it at volume. Paid search campaigns are running with fusion and quantum hardware teams on exactly that ladder.

### What does SuperMatics sell?

Our business combines platform software licenses, paid development programs, and licensing of materials and processes we own. Paid search campaigns are running with fusion and quantum hardware teams, and the platform is live with CAN Superconductors today. Four of the five paths to value sell into superconductors that are already in use.

### What is a search campaign?

A paid engagement with one brief: your device, what it has to do and where it stops. Say a grid team’s cable: 3 kA per phase at 77 K, a kilometer, a ceiling per kA·m. The platform sweeps the material, what it touches and how it is made against that brief and returns every route it can defend, ranked. Routes that clear come first, and routes that miss say by how much. A partner bench can prove the top one.

### How long does a superconductor R&D cycle take with SuperMatics?

An R&D cycle that took years now runs in weeks. A sweep runs in about a week today, iteration cycles included. The loop that used to cost a furnace run per answer costs a sweep. The leg that runs on partner-lab calendars is building and measuring, and that leg sets the schedule. So synthesis is moving in-house at Berkeley. We state the stages rather than one end-to-end number. The bench half depends on a queue we do not own, and a single figure would hide that.

### How do you know the predictions are any good?

We validate in both directions. Forward, our number goes on the record before the instrument runs. Backward, in a closed testing environment, settled superconductors go in with part of what is known withheld, and the platform recovers in minutes what the literature took decades to establish, scored against the published record alone. That is agreement with a century-deep record. It is not an accuracy figure, and we publish none: a measured Tc is not one number, since the onset, midpoint, zero-resistance and Meissner criteria differ by more than a kelvin.

### Has the physics been independently validated?

An effect consistent with a central MEL prediction was measured independently at Stanford and SLAC, with no involvement from SuperMatics. The measurement was published in Physical Review Letters in 2026. It concerns the physics behind our discovery research, not a test of the entire MEL framework or engineering platform. It does not establish that any proposed SuperMatics material superconducts.

### Who builds and measures what the search finds?

Our own lab work and partner benches, in fixed roles, and that division is deliberate. Synthesis runs at Berkeley, where it is moving in-house, and at industrial scale at CAN Superconductors. Crystals come from Brookhaven National Laboratory and the Walther-Meissner-Institut, and UIUC and Georgia Tech measure what comes back. We own the search, the program and the learning. Every result that comes home, hit or miss, trains the models. The benches that measure publish under their own names.

### What is the MEL framework?

The Modulated Electron Lattice framework informs our discovery research into new superconductors. It has been in development at Hyunsung TNC since 2006 and is licensed exclusively to SuperMatics. An independent experiment reported a result consistent with a central prediction. That does not validate the whole framework; our broader customer engineering work can use known superconductors.

### What is the highest-temperature superconductor?

At ambient pressure the durable record is Hg-1223 at 135 K, set in 1993, and it stood for 33 years. In 2026 a group in Houston reported 151 K in pressure-quenched Hg-1223: it is metastable, surviving about three days at 77 K, in flakes tens of microns across. Room temperature at ambient pressure has not been achieved.

### Why can’t a bigger lab just out-compute this?

The bottleneck is the theory. The large materials-AI efforts run on standard DFT, which is documented to fail for strongly interacting electrons. The platform uses models and experimental evidence to rank materials and manufacturing processes against a device brief. Customer programs can work with known superconductors. Separately, our discovery research draws on the Modulated Electron Lattice (MEL) framework to search for new materials. A founder reviews every record today; customer-operated software is planned.

### How is this different from a general materials-AI platform?

A general materials platform ranks materials. We return routes, scored against the brief you wrote. A route is the material, what it touches and how it is made. It carries the process window a fab can hold and the cost per meter, because a material a line cannot run is not an answer.

### Do you have a product or revenue today?

Yes. Paid search campaigns are running with fusion and quantum hardware teams, and the platform is live with CAN Superconductors today. Our team operates the software today; customer-operated enterprise access is planned. Revenue is early and unstated. Separately, our discovery research has five proposed materials in synthesis. Some are alloy-based candidates targeting high-temperature superconductivity. None has yet been independently established as superconducting. Provisional applications are on file; details remain confidential during the priority window.

### How big is the market?

Every sector we sell into already runs on a superconductor: MRI, transmission cable, quantum computers, maglev and particle accelerators among nineteen fields. None of them is a market we have to create, and every one is capped by how cold the conductor has to be kept. Systems built on superconductors are $10T+ a year by our own estimate; the sizing itself is in the deck.

### What intellectual property does SuperMatics hold?

SuperMatics holds 20+ patents issued and pending, with counsel by Wilson Sonsini, and an exclusive license to the MEL framework from Hyunsung TNC. The license is the part worth understanding. The right to build on the framework commercially is ours alone. The patent count is filings rather than inventions, and most are pending rather than granted.

### What are you not claiming?

We are not claiming a room-temperature superconductor, a shipped enterprise product, or a measured accuracy figure for predicted transition temperatures. What the platform returns are routes ranked against a constraint. What our own search sends to a bench is new and unnamed here. The closed test is retrospective, and an independent measurement of the underlying effect is not an independent measurement of a candidate. Our proposed materials have not yet been independently established as superconductors.

### How do I know whether my problem is a fit?

Send the device, what it has to do and where it stops. A founder reads it, and our first take comes back within a week, at no cost. The briefs we see most are a magnet holding current at field and temperature, a film surviving a fabrication process, a cable making a kilometer with joints in budget, and a conductor whose price the program cannot carry. Devices fail where two materials meet, so send the pairing with the constraint.

Source: https://supermatics.io/faq

## Open roles

### Founding CEO

Build the business around a platform for superconducting materials. Lead customer adoption, licensing partnerships, fundraising, and the team that takes SuperMatics forward.

- Location: Berkeley, CA or remote in the US
- Type: Full-time
- Compensation: $200-260k base · 3-7% equity
- Reports to: The board
- Responsibilities: Develop licensing relationships with manufacturers and turn technical opportunities into commercial agreements. Lead fundraising for the Series A and subsequent rounds, translating scientific and commercial progress into a clear investment case. Build and lead the executive team, with clear priorities across customer work, product development, and research. Represent SuperMatics with customers, investors, and the scientific community. Work with the founders to connect the platform to the markets it can serve.
- Qualifications: Experience scaling a technical company and leading the people responsible for its growth. A background in materials, chemistry, energy, quantum technology, or a related technical business. Direct experience raising a Series A or later financing round. Relationships in US or Korean industry and investing are a plus.
- Apply: https://supermatics.io/careers/apply/ceo

### Solutions Architect

Connect materials software to the laboratory. Turn proposed materials and processes into experimental programs that partner labs can execute and engineering teams can use.

- Location: Berkeley, CA or remote in the US
- Type: Full-time
- Compensation: $200-240k base · 1-2% equity
- Reports to: The founders
- Responsibilities: Translate model output into practical synthesis and characterization plans, with the conditions a laboratory needs to execute them. Build the workflow from computational results to partner experiments, keeping measurements and their provenance connected to the original brief. Design how partner laboratories contribute recipes, experimental conditions, and results to a customer program. Help engineering teams bring their requirements onto the platform and prepare the workflow for enterprise customers.
- Qualifications: A graduate degree in materials science, chemistry, physics, or computing. Hands-on experience running electronic-structure calculations and interpreting their limits. Fluency in Python and the ability to build dependable scientific workflows. Experience interpreting laboratory data, including measurement conditions, uncertainty, and conflicting results.
- Technical practice: You are hands-on with DFT, DFT+U, DMFT, or equivalent, and comfort reading STM/STS, XRD, or transport data is a plus.
- Apply: https://supermatics.io/careers/apply/solutions-architect

## Pages

- [Home](https://supermatics.io): Software, lab work and pilot manufacturing for superconducting devices. Find the material and process your device needs, then make and test it with us.
- [Platform](https://supermatics.io/platform): Explore superconducting materials and processes, compare the evidence, then make and test the route: lab work at Berkeley, pilot runs with partners.
- [Discovery](https://supermatics.io/discovery): The ambient-pressure record is 135 K and it stood for 33 years. This is the search we run against the band above it, and what happens to what it finds.
- [Company](https://supermatics.io/company): Meet the SuperMatics founders, officers and advisors bringing together superconducting physics, software, lab work and manufacturing in Berkeley and Suwon.
- [Investors](https://supermatics.io/investors): Software, lab work and pilot manufacturing for superconducting devices. Paid programs today, software licenses and owned materials next. Request the deck.
- [Reference](https://supermatics.io/reference): Ambient-pressure superconducting temperature records, MEL evidence and its limits, model testing, and applications in superconducting devices.
- [FAQ](https://supermatics.io/faq): How long a cycle takes, what we sell, who checks the work, whether it can predict a Tc, why the search needs AI at all, and what we are not claiming.
- [Glossary](https://supermatics.io/glossary): Twenty-five terms a superconductor program uses, defined in plain language: what each one means, and what it costs or decides for a team building a device.
- [Materials](https://supermatics.io/materials): The ten superconductors devices actually run on: what each one is, what it is made into, where it runs today, and the thing about it that ends programs.
- [Applications](https://supermatics.io/applications): Nineteen fields superconductivity has reached, from MRI and particle accelerators to fusion magnets. What each one gains, and what still stops it.
- [Notes](https://supermatics.io/notes): Occasional pieces on superconducting materials and the engineering around them: published records, what they establish, and what they do not.
- [Trust](https://supermatics.io/trust): The IP terms, the privacy controls the platform enforces, the ones the agreement does, and what nobody has certified yet. One page, so the two never disagree.
- [News](https://supermatics.io/news): What the company has done, newest first, and where the work has been reported.
- [Press](https://supermatics.io/press): Boilerplate written to be quoted, the checkable facts, and the logos, all on one page.
- [Contact](https://supermatics.io/contact): Research, investor, partnership, and press desks. A note here lands in a founder’s inbox, and the reply comes from one.
- Feed: https://supermatics.io/feed.xml
- News feed: https://supermatics.io/news.xml
- Careers: https://supermatics.io/company#careers

### Notes

- [The ambient-pressure superconductor record, 1911 to 2026](https://supermatics.io/notes/ambient-pressure-temperature-record): 2026-09-14

### Glossary terms

- [Superconductor](https://supermatics.io/glossary/superconductor)
- [Critical temperature (Tc)](https://supermatics.io/glossary/critical-temperature)
- [Kelvin (K)](https://supermatics.io/glossary/kelvin)
- [Ambient pressure](https://supermatics.io/glossary/ambient-pressure)
- [Meissner effect](https://supermatics.io/glossary/meissner-effect)
- [Cooper pair](https://supermatics.io/glossary/cooper-pair)
- [BCS theory](https://supermatics.io/glossary/bcs-theory)
- [Unconventional superconductor](https://supermatics.io/glossary/unconventional-superconductor)
- [High-temperature superconductor (HTS)](https://supermatics.io/glossary/high-temperature-superconductor)
- [Type II superconductor](https://supermatics.io/glossary/type-ii-superconductor)
- [Critical current density (Jc)](https://supermatics.io/glossary/critical-current-density)
- [Critical field](https://supermatics.io/glossary/critical-field)
- [Cost per kiloamp-meter ($/kA·m)](https://supermatics.io/glossary/cost-per-kiloamp-meter)
- [Cryocooler](https://supermatics.io/glossary/cryocooler)
- [Cuprate](https://supermatics.io/glossary/cuprate)
- [REBCO](https://supermatics.io/glossary/rebco)
- [Hydride superconductor](https://supermatics.io/glossary/hydride-superconductor)
- [Metastable](https://supermatics.io/glossary/metastable)
- [Doping](https://supermatics.io/glossary/doping)
- [Synthesis route](https://supermatics.io/glossary/synthesis-route)
- [Coated conductor](https://supermatics.io/glossary/coated-conductor)
- [Flux pinning](https://supermatics.io/glossary/flux-pinning)
- [Quench (in a magnet)](https://supermatics.io/glossary/quench)
- [Modulated Electron Lattice (MEL)](https://supermatics.io/glossary/modulated-electron-lattice)
- [Retrodiction](https://supermatics.io/glossary/retrodiction)

### Superconductors in service

- [NbTi](https://supermatics.io/materials/nbti): niobium-titanium, 9.2 K, wire
- [Nb₃Sn](https://supermatics.io/materials/nb3sn): niobium-tin, 18.3 K, wire
- [REBCO](https://supermatics.io/materials/rebco): rare-earth barium copper oxide, 92 K, tape
- [MgB₂](https://supermatics.io/materials/mgb2): magnesium diboride, 39 K, wire
- [Bi-2212](https://supermatics.io/materials/bi-2212): bismuth strontium calcium copper oxide, 85 K, wire
- [Nb](https://supermatics.io/materials/niobium): niobium, 9.3 K, film
- [Al](https://supermatics.io/materials/aluminum): aluminum, 1.2 K, film
- [TiN](https://supermatics.io/materials/titanium-nitride): titanium nitride, 4.5 K, film
- [NbN](https://supermatics.io/materials/nbn): niobium nitride, 16 K, film
- [NbTiN](https://supermatics.io/materials/nbtin): niobium titanium nitride, 15 K, film

### Fields

- [MRI](https://supermatics.io/applications/mri): In service today
- [Particle therapy](https://supermatics.io/applications/particle-therapy): In service today
- [NMR spectrometers](https://supermatics.io/applications/nmr-spectrometers): In service today
- [Biomagnetic sensing](https://supermatics.io/applications/biomagnetic-sensing): In service today
- [Transmission cable](https://supermatics.io/applications/transmission-cable): In service today
- [Fault-current limiters](https://supermatics.io/applications/fault-current-limiters): In service today
- [Wind generators](https://supermatics.io/applications/wind-generators): Built and demonstrated
- [Magnetic energy storage](https://supermatics.io/applications/magnetic-energy-storage): Built and demonstrated
- [Quantum computers](https://supermatics.io/applications/quantum-computers): Built and demonstrated
- [Photon counting](https://supermatics.io/applications/photon-counting): In service today
- [Superconducting logic](https://supermatics.io/applications/superconducting-logic): Waiting on a material
- [Telescope detectors](https://supermatics.io/applications/telescope-detectors): In service today
- [Maglev](https://supermatics.io/applications/maglev): Built and demonstrated
- [Electric aircraft](https://supermatics.io/applications/electric-aircraft): Waiting on a material
- [Ship propulsion](https://supermatics.io/applications/ship-propulsion): Built and demonstrated
- [Particle accelerators](https://supermatics.io/applications/particle-accelerators): In service today
- [Fusion magnets](https://supermatics.io/applications/fusion-magnets): Built and demonstrated
- [Magnetic separation](https://supermatics.io/applications/magnetic-separation): In service today
- [Induction heating](https://supermatics.io/applications/induction-heating): In service today

## Contact

- General: contact@supermatics.io
- Research: research@supermatics.io
- Investors: investors@supermatics.io
- Partnerships: partnerships@supermatics.io
- Careers: careers@supermatics.io

## Entity and citation

The company name is SuperMatics, one word with a capital S and M. Legal name: SuperMatics, Inc.

Official profiles:
- https://www.wikidata.org/wiki/Q141100455
- https://www.linkedin.com/company/supermatics
- https://www.youtube.com/channel/UCROzoeuVi6D6oWsmrbVbphQ

Use the linked page and original scientific publication when citing a claim. This summary supplements the website; it does not replace the evidence or its qualifications.

https://supermatics.io/llms.txt