Infleqtion, a quantum computing firm, announced on September 24 that it successfully entangled 30 logical qubits on its Sqale neutral-atom processor. The experiment, conducted in August, used 80 cesium atoms and approximately 1,000 physical operations. The output landed in the circuit's valid set roughly 25% of the time, a rate that is about 1,024 times higher than the uniform random expectation of 0.0244% (1 in 4,096). However, the company cautions that the results are preliminary and have not yet been peer-reviewed.
Technical Details of the Experiment
Each physical qubit in the Sqale system is a trapped neutral atom controlled by light. Infleqtion divided the 80 atoms into ten blocks of eight, with each block encoding three logical qubits, yielding a total of 30. The processor uses atom movement to achieve long-range connectivity, while focused optical control applies entangling operations in place. This design differs from superconducting machines, which rely on fabricated junctions and fixed wiring.
The test utilized an instantaneous quantum polynomial-time (IQP) sampling circuit. This type of circuit starts from a prepared superposition, applies commuting phase operations, and then measures the qubits. Infleqtion connected all ten code blocks and included four logical CCZ gates—three-qubit operations that extend the circuit beyond Clifford-only demonstrations.
Key Metrics and Engineering Innovation
The experiment's most notable engineering achievement is a new logical operation called double-CZ. The company says that GPT-5.6 Sol assisted in finding the circuit identity, which reduced the number of physical two-qubit gates required for each logical entanglement from eight to four. This cut the local cost in half, potentially reducing error accumulation. The compiled experiment used 136 physical two-qubit gates across the full circuit, with all-to-all logical connectivity maintained through atom movement.
The [[8,3,2]] code notation summarizes the trade-off: eight physical qubits encode three logical qubits with a code distance of two. This high encoding rate is economical but cannot correct an arbitrary unknown error; it can only detect one error or recover a known atom loss.
Comparison with Previous Work
This achievement builds on earlier Sqale work. An April 2026 preprint reported 8- and 12-logical-qubit circuits with two-to-fourfold lower error rates than unencoded versions. The new run scales that architecture to a larger connected state. However, comparing logical-qubit counts across platforms is not straightforward. A peer-reviewed Harvard-led neutral-atom experiment previously entangled 48 logical qubits using up to 280 physical atoms, 228 logical two-qubit gates, and 48 logical CCZ gates. That work also demonstrated error-detected performance above physical-qubit baselines.
Infleqtion's claim is narrower: it is the first to achieve 30 logical qubits on a commercial Sqale system. The notable figure is the 80-to-30 physical-to-logical ratio, which may matter if later tests show durable error suppression. An independent Quantum Data assessment echoed caution, noting that logical error rates and peer-reviewed methods have not yet been published.
Limitations and Future Steps
The test does not establish error correction. The blocks were prepared with a distance-three code, but the downstream circuit used a distance-two encoding. Distance two can detect an arbitrary single error but cannot generally correct it. Infleqtion used known atom loss information to reconstruct some lost measurements, reportedly quadrupling the number of usable shots, but also increasing error rates—a throughput-versus-accuracy trade-off.
Membership in the allowed set is a coarse test; it does not show whether valid outputs appeared with correct relative probabilities. A noisy device could overproduce a smaller group and still register hits. The company says a full technical paper will follow within weeks, which should include a fuller distribution metric.
For investors and industry watchers, this development underscores the rapid progress in neutral-atom quantum computing, but also the gap between laboratory demonstrations and practical error-corrected systems. As quantum hardware advances, the ability to scale logical qubits while maintaining low error rates will be critical for commercial viability.
