Roadmap
A measured path toward fault-tolerant quantum computing
Where the neutral-atom platform stands, and the development goals we are working toward. Demonstrated results and targets are kept apart on purpose.
Talk to the teamMilestones
2026
Demonstrated
- 50+ physical-qubit neutral-atom platform launched
- AI decoder benchmarked on real hardware data with NVIDIA CUDA-Q QEC
- LayerQake validated on three IBM quantum processors
2027
Target
- Thousands of physical qubits by end of H1 2027
- AI-driven decoder integrated with the control stack
- Toward real-time logical-qubit operations
2028
Development goal
- Error-corrected logical qubits at scale
- Continued growth in physical qubit count
- Higher decoder throughput
2030
Development goal
- Working toward fault-tolerant quantum computing
- A modular platform with integrated error correction
- Error correction, not qubit count, defines the machine
Milestones beyond 2026 are development goals, not commitments. They depend on platform scaling, real-time decoder integration and error rates no roadmap can guarantee.
From physical to logical qubits
Many physical qubits, fewer logical ones. Atoms held in an optical tweezer array are grouped and error-corrected into logical qubits — the unit that long, reliable programs run on. Illustration only.
Why neutral atoms
Naturally uniform
Every atom of an isotope is identical — no two qubits need matching.
Potential for long coherence
Neutral atoms couple weakly to their surroundings; the platform is designed for extended coherence times.
Reconfigurable arrays
Optical tweezers rearrange the register dynamically — the layout is software, not silicon.
None of this removes the need for error correction: atoms are still lost and gates still err. That is why the decoder sits at the centre of the plan.
Build the next milestone with us
Hardware teams, labs and partners: tell us where you are on your own path.
Talk to the team