A quantum photonic processor built on a six-mode integrated glass circuit has completed eight months of operation in low Earth orbit, marking the first time a quantum computer has run in space. The device launched on a SpaceX Falcon 9 rocket on June 23, 2025, and continued to generate, manipulate and detect photon pairs despite losing half its single-photon detectors to launch stresses.

The system comes from a team of physicists led by Philip Walther at the University of Vienna. Their paper, posted to the arXiv preprint server (Simon Steiner et al., "In-orbit operation of a programmable quantum photonic processor," DOI: 10.48550/arxiv.2609.25248), reports a working demonstration of programmable quantum photonics outside the laboratory — a milestone for a hardware class that vendors typically run in climate-controlled racks.

Why orbit needs onboard compute

The commercial driver is bandwidth. Satellites collecting Earth-observation and sensing data face bottlenecks when they downlink raw measurements to ground stations. Processing that data in situ — filtering, compressing or running inference before transmission — would cut the volume of data crossing the link. Quantum photonic processors are one candidate for that onboard processing role, provided they survive radiation, thermal cycling and vacuum.

The Vienna hardware addresses the survival question directly. The flight package combined a photon-pair source, the six-mode glass chip, single-photon detectors and control electronics — a compact optical architecture with no cryogenic requirements, which is what makes orbital deployment plausible at all.

How the processor works

Operation follows a standard photonic-computing recipe. A laser shines through a crystal to produce entangled photon pairs. Those photons pass through the glass circuit, where microscopic heaters steer their paths and control interference. Detectors at the output record where each photon emerges, reading out the result of the programmed operation.

The mission hit its core goals. Over the first eight months in orbit, the system executed several different programmed operations and successfully observed two-photon interference — the quantum-mechanical effect that underpins photonic quantum computation — under unshielded space conditions.

The hardware did not emerge intact. Half the photon detectors, three of six, failed after launch. The team tuned the remaining detectors and pressed on, a partial loss that itself carries engineering weight for anyone designing radiation-tolerant detector arrays for future flight quantum hardware.

A first step, not a product

The researchers are explicit about the gap between demonstration and deployment. Running useful quantum-assisted data processing on a satellite requires more than interference in a glass chip.

"The next step is to close the loop between sensor and processor, encoding Earth-observation data directly into the unitary programmed on the circuit," the authors write in the paper.

Stability is the other constraint. "What remains is to keep it stable over the long acquisition times that an inference task demands, as the gradual degradation of the components reduces the coincidence rate," the authors note. In other words, radiation-driven component drift erodes the photon-pair measurement rates that any real inference workload would depend on, and long-duration missions will need compensation strategies.

Context for the industry

For the photonics supply chain, the result is an early but concrete data point: integrated glass photonic circuits, laser-driven pair sources and single-photon detectors can be qualified for launch and operated in orbit, at least for months. That matters because most discussions of space-based quantum systems have centered on quantum key distribution, not on programmable processors. Walther's group has moved the conversation from quantum communications hardware to quantum compute hardware in orbit.

The economics remain distant. No revenue figure, capacity plan or product family attaches to this mission — it is a university experiment on a rideshare-class launch, not a commercial deployment. But the direction is clear: if satellite operators want onboard processing to relieve downlink constraints, quantum photonics now has flight heritage to point to, and component-level radiation tolerance has become the next engineering problem to solve.