The Signal
Rosetta Audit Finds Quantum Roadmaps Slip Up to Three Years
Rosetta Quantum compared nine dated commitments across seven quantum companies with their latest public statements. Four targets moved later, including Rigetti’s 1,000-qubit system, which shifted from 2024 to 2027.
Quantum.world Team · Written with AI · fact-checked ·
Rigetti’s 2024 target now points to 2027
Quantum companies have set ambitious delivery dates; a new audit finds that some of those deadlines have moved by years, while others have held or shifted earlier. In an assessment dated October 29, Rosetta Quantum compared nine dated commitments across seven organizations with each company’s latest public statement about the same milestone. Four commitments moved later by 12 to 36 months, one moved a year earlier, and two were delivered on schedule. Two fault-tolerance targets cannot yet be judged because their dates fall in 2029.
The largest measured delay is Rigetti’s 1,000-qubit system. The company’s October 2021 roadmap placed it in 2024; public statements later moved the target to late 2025 and then to 2027. Rosetta calculates a 36-month slip from the original date. The company’s 4,000-qubit target, once set for 2026, was pushed to 2027 or later and no longer had a date in the latest statement the audit examined.
PsiQuantum’s Brisbane schedule carries a reporting caveat
PsiQuantum set an end-of-2027 target for its Brisbane site in connection with a 2024 funding announcement. Australian reporting in December 2025 put construction about a year behind schedule and the utility-scale timeline closer to 2030. Rosetta counts that as an approximate 24-month delay, but stresses that the revised date comes from press coverage, not a PsiQuantum filing.
Nord Quantique also has two public timelines that do not align neatly. Its roadmap targets 100 logical qubits by 2028, while its own press language describes a path to a machine with more than 100 logical qubits by 2029. Rosetta records a 12-month gap between those statements, while flagging that the company has not reconciled them.
IBM and Quantinuum offer counterexamples
The audit does not treat every forecast as a miss. IBM delivered the early rungs of its processor roadmap on schedule, culminating in the 1,121-qubit Condor announced in December 2023. The first milestones on its later fault-tolerance plan also arrived in 2025: IBM announced its Loon test chip and 120-qubit Nighthawk that November.
Quantinuum’s public target moved in the opposite direction. A September 2024 roadmap set universal, fully fault-tolerant quantum computing for 2030; the later Apollo target is 2029. Rosetta counts that as a 12-month move earlier, not as a completed milestone: the date cannot be tested until 2029.
IonQ’s result depends on whose metric counts
IonQ’s roadmap called for broad quantum advantage by 2025, and its Tempo system was reported at AQ 64 that September. Rosetta marks that commitment on schedule only under IonQ’s own metric. The audit says AQ is defined and reported by the company, and that no neutral party has adjudicated whether the milestone amounts to broad quantum advantage.
That distinction matters to the audit’s headline numbers. Rosetta counts only pairs with dated documents at both ends, including a dated commitment and a later statement or delivery. It does not assign a numerical delay when a milestone loses its date, and it says public statements cannot establish what is happening inside a company’s engineering program.
The 2029 targets are still ahead
Rosetta describes its nine comparisons as a sample, not a census. Undated roadmaps cannot be scored, so the measured slips do not establish an industry-wide rate; the report says that missing dates may hide delays. Google’s 2029 target for a useful, error-corrected machine remains unassessed in the audit because Rosetta found no public revision, which does not show that the internal schedule is on track.
Several prominent fault-tolerance goals now point to 2029, including IBM’s Starling and Quantinuum’s Apollo. Rosetta says those commitments will only become testable when the dates arrive, and plans to use roadmap credibility in its scheduled April 2027 update to a pre-registered Q-Day estimate. Until then, the audit’s longest scored slip is Rigetti’s 1,000-qubit system: from 2024 to 2027.