Classical computing protects information by copying it. Quantum computing can’t do that, so the path to real-world impact runs through quantum error correction: detecting mistakes without destroying the quantum state.

A fresh result from ETH Zurich (with collaborators including PSI) points to a practical way forward: keep error correction running while logical operations happen, instead of treating “protect” and “compute” as two separate phases.

What’s new here (without the heavy math)

Most recent milestones in error correction have focused on preserving one logical qubit. Important, but not enough. Real algorithms need you to operate on logical qubits: entangle them, move information around, and build larger computations.

That’s where lattice surgery comes in: a technique designed for the leading “surface code” roadmap, where qubits live on a 2D grid and only interact with neighbors.

In this work, the team demonstrates a core primitive of lattice surgery on a 17-qubit superconducting device: they split one encoded (surface-code) logical qubit into two encoded repetition-code qubits, enabling a protected, logical two-qubit operation (a foundational step toward logical gates).

Why it matters

Think of it as moving from “we can keep one logical qubit alive” to “we can start doing useful operations on protected qubits.” It’s still a proof of concept (and not yet fully fault-tolerant against every error type), but it’s the kind of engineering milestone that turns quantum from a lab demo into a scalable system design.

What Quantum Circle is watching next

For Belgium’s quantum ecosystem, this is a signal worth tracking across our workgroups:

  • Technology & Research: surface-code architectures and logical operations are converging into repeatable building blocks.
  • Business & Ecosystems: these primitives shape hardware roadmaps, and where suppliers, talent, and use cases can align early.
  • Society: the closer we get to fault tolerance, the more urgent it becomes to prepare skills, governance, and adoption pathways.

Read on

Care for the deep technical details? Go straight to the source and read the paper

Related posts