Where the tape goes
High-temperature superconductor applications
Six markets buy superconducting conductor today, most of them drawing on the same tape from a small number of factories. What separates them is how much they need, what they will pay, and whether anything is being built.
The supply picture
Roughly 15 manufacturers worldwide produce high-temperature superconductor (HTS) tape, with a combined annual capacity above 5,000 km of 12 mm-equivalent tape. Four of them, all using pulsed laser deposition, account for more than half of world output between them. Tape sells for about $15 to $30 per metre, and its cost still has to fall: to $50 per kilo-ampere-metre in the near term and $10 to $20 in the long run (Superconductivity, 2025). The targets use a different unit from the price: they divide cost by the current the tape carries, so that tapes of different performance can be compared (how the tape is made explains it).
Those three facts explain most of what follows. Supply is concentrated, capacity is small in absolute terms, and price still has a long way to fall. Applications that can absorb expensive conductor are being built now; the rest are waiting on the cost curve.
At a glance
| Application | Stage | Tape demand | Buyers | What decides it |
|---|---|---|---|---|
| Fusion magnets | Demonstration magnets and first machines | Largest, and the reason capacity is expanding | Private fusion developers | Tape capacity keeping pace with reactor timelines |
| Data centres | Demonstrations, with launches targeted for 2027 to 2030 | Small today, the fastest-growing | Data centre operators and their cable suppliers | Power density inside AI campuses |
| Grid transmission and distribution | In service at a handful of city sites | Steady, one project at a time | Electricity utilities | Public funding, and cities with no room for new routes |
| Rail traction power | First installations, on the French network | Small, and will stay small | Rail network operators | Substation capacity in city centres |
| Naval and defence | In service on US Navy ships | Small volume, high value per metre | Navies | Defence procurement cycles, not tape price |
| Particle accelerators | Established | Little HTS tape: MgB₂ won the power links | Public research laboratories | Laboratory procurement cycles |
Fusion magnets
The largest single pull on tape supply
Fusion is the reason the tape industry is expanding. A tokamak's power output rises steeply with magnetic field strength, roughly with its fourth power, so a stronger magnet buys a much smaller and cheaper machine. REBCO tape carries very high currents in fields beyond anything low-temperature superconductors survive, which is what makes a compact, high-field reactor arguable at all.
The volumes are what matter commercially. Faraday Factory Japan alone had delivered a cumulative 5,000 km of tape to fusion developers by September 2024, which is roughly the whole world's current annual output. Fujikura has committed two rounds of capacity investment in two years and names fusion as the reason. When a supplier expands, read the press release for which end market it blames: at present it is usually this one.
For an investor the consequence is that fusion timelines and tape capacity are now the same question. A reactor programme cannot outrun its conductor supply, and conductor suppliers are pricing accordingly.
Recent coverage
Data centres
The fastest-growing new application
Data centre campuses need to move tens or hundreds of megawatts across short distances inside a constrained site. That is an unusually good fit for a superconductor: very high current at modest voltage, over a run measured in hundreds of metres rather than tens of kilometres, where the cost of the cryogenic plant is spread over a large power flow and the space saved has real value.
This application did not exist in the news at all before mid-2025. Since then there has been a steady run of announcements, including a Korean utility and cable maker pairing on an in-campus superconducting grid, a megawatt-scale demonstration in a simulated data centre, an Italian project running a cable inside a working research data centre, and a Japanese wire maker and cable builder agreeing to develop in-building distribution for the end of the decade.
None of it is deployed at scale yet. VEIR is aiming for a commercial launch in 2027, and SWCC and Faraday Factory Japan for working use by fiscal 2030. What makes it worth watching is that the customer has money, an urgent constraint, and no strong attachment to conventional copper.
In the cable table: Ishikari DC cable project
Grid transmission and distribution
The oldest application, and the deepest project record
Putting superconducting cable into a public electricity grid is the application with the longest history and the most completed hardware. A superconducting circuit moves several times the power of a conventional cable through the same trench, which is worth most in dense cities where digging a new route is expensive or simply impossible, and in substations where the alternative is a new building.
This is what our project table covers in detail: cables in Essen, Chicago, Shanghai, Shenzhen, Yongin, St Petersburg and Long Island, each with its source, alongside planned links such as the 110 kV SuperLink in Munich. Fault current limiters belong here too, because they do something no conventional device can, passively limiting a fault and then resetting themselves.
The honest commercial read is that grid deployment has been slow and mostly grant-funded for two decades. It is a real market with real installed hardware, but it is no longer the one setting the pace.
In the cable table: AmpaCity, SuperLink, Shanghai Xuhui 35 kV cable
Rail traction power
Small, but an unusually good technical fit
DC railway networks move very large currents at low voltage, which is precisely a superconductor's preferred regime, and substation capacity is what limits how many trains can run. Reinforcing the feed with a superconducting cable can add capacity without building a new substation, which in a city centre may be the only option available at any price.
The Paris Montparnasse SuperRail installation feeds the 1,500 V DC traction supply at 3,500 A per cable, a world first. Separately, Nexans and SNCF Réseau put a superconducting fault current limiter on the Belfort to Delle line, again a first for a rail power supply.
Volumes here are small and will stay small. The value is as a reference: rail operators are conservative buyers, and a working installation on a national network is a credential that transfers to other infrastructure customers.
In the cable table: SuperRail, Paris Montparnasse
Recent coverage
Naval and defence
Small volume, high value per metre
Warships have to manage their magnetic signature, and the coils that do it are a natural fit for superconductors: the same job done with far less weight and electrical load, both of which are tightly budgeted aboard a ship. Degaussing is the established use; superconducting propulsion motors and generators have been demonstrated but not fielded at scale.
AMSC won roughly $75 million for its first allied-navy Ship Protection System order, for the Royal Canadian Navy's new surface combatants, following earlier deployment on United States Navy amphibious ships.
For an analyst this segment behaves differently from the rest: procurement cycles are long, contract values are disclosed, margins are defence margins, and the customer is insensitive to conductor price. It is the one application where cost per kilo-ampere-metre is not the deciding variable.
Recent coverage
Particle accelerators
Established and steady, but not a tape market
Accelerator laboratories were using superconducting magnets long before anyone spoke about high-temperature materials, and they remain a reliable consumer of conductor. The newer role is in power transfer rather than the magnets themselves, moving very large currents from power supplies to magnets through flexible cryogenic links. That work has gone to magnesium diboride (MgB₂), which superconducts at 39 kelvin, well below the liquid nitrogen line that defines a high-temperature superconductor, and which is cheap to draw into wire.
CERN's High-Luminosity LHC superconducting links are the reference case: flexible cryostats each bundling 19 MgB₂ cables, rated for a combined 120,000 amps, with 1,450 km of MgB₂ wire produced for them by October 2023. That is a large conductor order by any measure, and none of it was HTS tape.
The commercial character is unusual. Buyers are public laboratories with long procurement cycles and published requirements, which makes demand forecastable but slow-moving and largely independent of the fusion cycle.
In the cable table: HL-LHC superconducting links
Recent coverage
Keep exploring
For the installations themselves, with lengths, ratings, suppliers and sources, see the cable projects table. For how the conductor is made and why it costs what it does, read how REBCO superconductor tape is made. For what the finished cable actually looks like inside, explore the cable anatomy cutaway. What this site covers and what it leaves out is set out on the about page.