IonQ démontrera-t-il publiquement au moins 12 qubits logiques avant le 31 mars 2027 ?

• Ce qui fait monter. IonQ's official roadmap targeted 12 logical qubits for 2026, meaning a delivery before 31 March 2027 requires only minor or zero slippage into Q1 2027 [tier 1]
• Ce qui fait monter. IonQ advanced to DARPA QBI Stage C under a $300M agreement, indicating external validation of its fault-tolerant quantum computing trajectory [tier 1, article 9]
• Ce qui fait monter. Recent technical focus on qLDPC error correction and interconnects demonstrates active engineering focus on fault-tolerant logical scaling [tier 1, articles 5, 11]
• Ce qui fait descendre. As of mid-October 2026, no public demonstration of 12 logical qubits has materialized, leaving barely 5.5 months before the 31 March 2027 deadline [tier 1]
• Ce qui fait descendre. Demonstrating full active error correction across 12 logical qubits is an exceptionally high technical hurdle, and quantum hardware roadmaps frequently slip by multiple quarters [tier 1]
• Ce qui fait descendre. Recent publications focus heavily on theoretical or scheduled qLDPC speedup architectures, corporate M&A, and financing rather than verified experimental active error correction runs [tier 1, articles 2, 3, 11]
• Ce qui ferait le plus bouger ce chiffre. Whether the Superion 256 platform is already executing active syndrome measurement cycles in the lab or if active error correction testing remains in preliminary phases.

• Ce qui fait monter. IonQ has begun deploying its 'Superion 256' platform to partners like FIU and NVIDIA, providing the necessary physical qubit overhead (256 qubits) to support 12 logical qubits [9, 11].
• Ce qui fait monter. The company has advanced to Stage C of DARPA’s Quantum Benchmarking Initiative, indicating that their technical approach to fault tolerance has passed rigorous external verification [5, 6].
• Ce qui fait monter. IonQ is actively publishing research on qLDPC codes and scheduled error correction, which are specifically designed to enable logical qubits with lower physical overhead [7].
• Ce qui fait monter. The 2026 roadmap explicitly targets 12 logical qubits, and the company is currently in the final quarter of that target year [Starting State].
• Ce qui fait descendre. Despite being in Q4 2026, IonQ has not yet publicly demonstrated the 12 logical qubit milestone promised for this year [Starting State].
• Ce qui fait descendre. The transition from physical qubits to active error correction on a dozen logical qubits is a major technical hurdle that has historically seen delays across the industry.
• Ce qui fait descendre. Recent major announcements have focused on interconnect speeds and system deployments rather than a breakthrough in logical qubit counts [1, 2].
• Ce qui ferait le plus bouger ce chiffre. Whether the gate fidelities on the Superion 256 platform have reached the 99.99% threshold required by their roadmap to effectively implement the 12-logical-qubit error correction scheme.

• Ce qui fait monter. IonQ's official roadmap targeted 12 logical qubits for 2026, putting the target window directly within or immediately preceding the deadline [tier 1].
• Ce qui fait monter. IonQ has actively unveiled Superion 256 hardware partnerships and published progress on qLDPC error correction and CPU-powered decoders in September-October 2026 [tier 1, articles 4, 6, 8, 9].
• Ce qui fait monter. Advancement to DARPA's Quantum Benchmarking Initiative Stage C suggests third-party technical validation of their fault-tolerance path [tier 1, article 3].
• Ce qui fait descendre. As of mid-October 2026, no demonstration of 12 logical qubits has been published or preprinted, leaving fewer than six months before the 31 March 2027 deadline [tier 1].
• Ce qui fait descendre. Hardware error-correction demonstrations at scale are prone to slippage, and recent releases highlight architecture and interconnect progress rather than verified active logical qubit operations [tier 1, articles 0, 1, 4].
• Ce qui ferait le plus bouger ce chiffre. Whether the Superion 256 platform can operate with sufficiently low physical error rates to implement active qLDPC decoding on 12 logical qubits in experiment rather than simulation before Q1 2027 ends.

• Ce qui fait monter. IonQ has announced its 'Superion 256' platform featuring 256 physical qubits, which provides the necessary physical qubit overhead to support 12 logical qubits [tier 1, Article 4, 5].
• Ce qui fait monter. IonQ recently demonstrated a real-time, CPU-powered quantum error correction decoder, solving a major bottleneck for active error correction [tier 1, Article 6].
• Ce qui fait monter. IonQ has advanced to Stage C of DARPA's Quantum Benchmarking Initiative, indicating rigorous external validation of their scaling roadmap [tier 1, Article 1].
• Ce qui fait descendre. No public demonstration of 12 logical qubits has been published yet, despite the original 2026 roadmap target, leaving a tight window of under 6 months [tier 1, Starting State].
• Ce qui fait descendre. Implementing active error correction across 12 logical qubits is an incredibly complex engineering challenge that may experience integration delays [tier 1, Starting State].
• Ce qui ferait le plus bouger ce chiffre. Whether the newly announced Superion 256 systems can successfully execute active error correction protocols to sustain 12 logical qubits in a published demonstration before the March 2027 deadline.

• Ce qui fait monter. Tier‑4 article 0 reports a new qLDPC method that could speed up logical gates, suggesting progress toward error‑corrected qubits [4]
• Ce qui fait monter. Tier‑4 article 1 claims a real‑time error‑correction breakthrough, indicating possible nearing of logical‑qubit demonstrations [4]
• Ce qui fait monter. Tier‑4 article 2 describes an end‑to‑end real‑time quantum error decoder, a component needed for logical qubits [4]
• Ce qui fait monter. Tier‑4 article 3 mentions targeting the error‑correction bottleneck with a single‑CPU decoder, implying hardware readiness [4]
• Ce qui fait monter. Tier‑4 article 4 reports a decoder running 31.5 million operations, showing scaling of error‑correction infrastructure [4]
• Ce qui fait descendre. Hand‑verified tier‑1 roadmap shows the 12‑logical‑qubit goal for 2026 but no demonstration has been found three‑quarters through 2026, indicating a delay [1]
• Ce qui fait descendre. All recent reports are tier‑4 and uncorroborated by tier‑1 or tier‑2 sources, so they are weak signals and do not confirm a shipped demonstration [4]
• Ce qui ferait le plus bouger ce chiffre. A confirmed, peer‑reviewed paper or official IonQ release documenting 12 logical qubits before 31 Mar 2027 would dramatically raise the estimate.

Raisons non conservées pour cette révision (l'historique détaillé n'est gardé que depuis peu).