Sixty qubits. One fifth the power draw. Not a single laser anywhere in the machine. That is the pitch behind JION, the trapped-ion quantum computer that Forschungszentrum Jülich switched on this month inside one of Europe’s busiest supercomputing centers. Researchers there worked with the German startup eleQtron to build a machine around a control method that trades laser optics for plain microwaves, and the swap could make laser-free trapped-ion computing the more practical route to bigger, cheaper quantum machines.
How Laser-Free Trapped-Ion Computing Works
Most trapped-ion machines control each qubit with a laser beam aimed at a single ion suspended in a vacuum trap. That works, but it demands tight alignment, vibration isolation, and an optics rig that gets more complicated every time you add qubits. eleQtron’s answer is a technique called Magnetic Gradient Induced Coupling, or MAGIC. A static magnetic field gradient runs across a chain of trapped ytterbium ions, so each one ends up resonating at a slightly different frequency. From there, ordinary microwave pulses tuned to 12.64 gigahertz can address one ion at a time, the way a radio dial tunes into a single station out of a crowded band. Skip the lasers, and the entire optical bench disappears with them.
A Modest Fidelity Trade for a Bigger Manufacturing Win
JION runs 60 qubits and hits roughly 99.7 percent fidelity on its Bell-state benchmark. That trails the 99.9 percent Quantinuum claims and the 99.99 percent IonQ has posted with laser-controlled ions, so this is not a raw-performance record. What eleQtron gained instead is a control system built from standard RF electronics rather than specialized optics, plus a partnership with chipmaker Infineon aimed at series production on ordinary semiconductor lines. Trade a little precision now, in other words, for a design a fab can actually stamp out by the hundreds later.
Wired Straight Into a Supercomputer
Jülich did not stop at installing JION in a lab. The machine plugs into JUNIQ, the center’s unified quantum infrastructure, which links quantum processors directly to the Jülich Supercomputing Centre’s HPC clusters. Scientists can now hand off individual calculation steps in chemistry, materials research, logistics optimization, and machine learning to JION mid-run, testing hybrid quantum-classical workflows without moving data between separate facilities.
It is a similar instinct to the one behind Diraq’s silicon-spin quantum computer moving into a commercial data center earlier this month: a quantum processor sitting alone in a physics lab helps nobody, but one wired into infrastructure people already use starts pulling its weight. Not every lab is chasing scale the same way, either. MIT’s new qubit architecture aims to make superconducting quantum computers ten times faster by refining the qubit itself, while JION bets that simplifying the control electronics around the qubit is the faster route to thousands of them.
Picture Amara, a 34-year-old chemistry postdoc in Lagos who collaborates remotely with a materials-science group in Germany. She will never fly to Jülich to babysit a laser table, and now she does not have to. A quantum processor built from standard RF electronics that a semiconductor fab can mass-produce is also one a university IT department can keep running, not just a laser physicist. That kind of unglamorous reliability is what turns a research toy into a shared resource other scientists can actually book time on.
eleQtron CEO Jan Henrik Leisse described JION as the integration of “trapped-ion and MAGIC technology into one of Europe’s leading user infrastructures for quantum computing.” Sixty qubits will not break any encryption or model a new drug on their own. But a laser-free design that a semiconductor fab can manufacture, sitting inside a live supercomputing center instead of an isolated lab, gives trapped-ion quantum computing a route to scale that does not depend on chasing the next fraction of a percent in fidelity.