IBM publishes a roadmap for quantum computing research, estimating it will make a processor with 1,000+ qubits by 2023 and debut a 127-qubit processor next year
Context & Ripple Effects
The roadmap formalizes what IBM had been building since it put a 20-qubit machine on the cloud and showed a 50-qubit prototype in 2017: quantum stops being an annual demo and becomes a dated delivery schedule, with a 127-qubit chip promised within a year and 1,000+ qubits by 2023.
The later record shows the schedule held — Osprey's 433 qubits landed on top of the 127-qubit Eagle — and the ambition kept compounding toward a 100,000-qubit system within ten years, which is why this 2020 document reads as the moment IBM turned quantum into a public commitment its rivals would be measured against.
First-order effects
- IBM binds itself to named qubit milestones — 127 next year, 1,000+ by 2023 — converting its quantum program from laboratory announcements into a checkable cadence that customers on its cloud service can plan around.
Second-order effects
- Competitors like Google get a public benchmark to chase, pushing the field into roadmap-versus-roadmap competition where each announced qubit count pressures the rival's next reveal.
Third-order effects
- If the cadence holds, the binding constraint shifts from raw qubit counts to error correction at scale — the gap IBM and Google later frame as the hurdle between hundreds and millions of qubits, and the reason IBM eventually commits more than $10B toward a large-scale, error-free machine by 2029.
The trend: Quantum computing is moving from research demonstrations to dated, capital-intensive engineering roadmaps racing toward fault-tolerant machines by the end of the decade.