What was shown
Quantum bits are fragile. The researchers grouped several physical qubits into one more reliable unit, called a logical qubit. Think of a bank teller who counts a cash drawer twice and sets aside any batch where the two counts disagree. Here, checks detected errors, and runs that flagged one were discarded.
On a real machine of up to 280 atoms held in place by laser light, they ran test programs on 48 such units together. The programs used 228 gates acting on two units and 48 gates acting on three units at once. The output looks random, but its quality can be scored. The score reached roughly 0.1, up to about an order of magnitude above earlier runs on unprotected qubits of similar complexity. Growing the code from size 3 to size 7 also made a two-unit gate better.
“We obtain an XEB of ≈0.1 for 48 logical qubits and hundreds of nonlocal logical entangling gates, up to roughly an order of magnitude higher than previous physical qubit implementations of similar complexity [18, 47], showing the benefits of a logical encoding for this application.”
— Section "Complex logical circuits using 3D codes" (Figs. 5e,f), in: Logical quantum processor based on reconfigurable atom arrays1
- How it was done: run on a physical quantum processor
- Qubits: Up to 280 physical qubits, made of individual atoms held by laser light, were used to encode up to 48 logical qubits.
- Error correction: Errors were detected and flawed runs discarded; correcting errors again and again while a program runs was not done.
- Compared against classical computing: No advantage over classical computers is claimed; the circuits stay easy to simulate, and results were compared with earlier physical qubit machines.
Who it affects
Nobody can use this today. The circuits are benchmarks with no commercial task behind them, and an ordinary computer can still simulate them. The result concerns anyone who expects quantum computers to run useful jobs one day, because such jobs need reliable qubits in large numbers. For those readers nothing changes yet. What the paper adds is evidence that dozens of error-protected qubits can be operated together on real hardware.
Quantum computing vendors and their customers: McKinsey & Company estimates $43B to $71B by 2035 — quantum computing market size (vendor side) (Quantum Technology Monitor, April 2026, p. 8).
This figure is an estimate, not a measurement: McKinsey & Company is a consultancy with its own quantum advisory business. It covers a whole use case, not the result of this paper.
What is still missing
The protection here was limited. The machine detected errors and threw away flawed runs, and it used small codes. Correcting errors again and again while a program runs, which the authors name as a key next step, was not done. Lower error rates in two-qubit gates and far more physical qubits than 280 are also missing. The circuits were sampling and scrambling benchmarks, not an industrial task, and a classical method calculates each result in about 2 seconds.
Who is working on it
The paper (December 2023) lists its authors' affiliations as: Department of Physics, Harvard University, Cambridge, MA 02138, USA; QuEra Computing Inc., Boston, MA 02135, USA; Joint Center for Quantum Information and Computer Science, NIST/University of Maryland, College Park, Maryland 20742, USA; John A. Paulson School of Engineering and Applied Sciences, Harvard University, Cambridge, MA 02138, USA; and 1 further institution. That is who signed this work — it is not a ranking of who is ahead.
It strengthens rearrangeable arrays of neutral atoms, built in zones, with gates applied to whole code blocks and control at the level of the encoded qubits, as a route to error-corrected quantum computing.
How this brief was made
This brief was written by software, not by a person: a language model read the full text of the paper, and code then checked every quotation, number and name in this brief against that text before it was published. The rules are set out under Method. Responsible for the rules and for this page: Georg Singer. The paper is peer-reviewed, published in Nature (2023). Georg Singer holds no shares in any company or institution named in this brief (as of October 2026).