Microsoft Majorana 2 Aims for Practical Quantum by 2029
Key insights
- Majorana 2 qubits are 1,000 times more reliable than the prior generation, with mean lifetimes of 20 seconds and some reaching approximately one minute.
- Lead replaces aluminum as the superconductor; indium arsenide antimonide updates the semiconductor region for a more stable topological phase.
- Microsoft's Discovery agentic AI system co-designed Majorana 2, marking an early AI-hardware design milestone at a major technology company.
Why this matters
Microsoft's 2029 target for practical quantum computing gives enterprise customers a concrete planning horizon tied to a major cloud vendor's public commitment for the first time. The use of Discovery agentic AI to help design Majorana 2 sets a precedent for AI systems contributing directly to hardware R&D, a pattern that could compress development cycles well beyond quantum. For technical leaders, the combination of a 1,000x qubit reliability improvement and a named utility date is specific enough to drive real quantum readiness planning at the infrastructure and software stack level.
Summary
Microsoft revealed Majorana 2 at Build 2026 in San Francisco, claiming a practical quantum computer by 2029.
Lead replaces aluminum as the superconductor; the semiconductor region shifts to indium arsenide and indium arsenide antimonide. Chetan Nayak, Microsoft's technical fellow and VP of quantum hardware, says this creates a more stable topological phase that shields qubits from cosmic disturbances. Qubits are 1,000 times more reliable than Majorana 1, with a mean lifetime of 20 seconds.
Essentially: (Microsoft, Discovery AI) Majorana 2 was co-designed with an agentic AI platform.
- Some qubits lasted approximately one minute, a 1,000x improvement over Majorana 1
- Lead superconductor; indium arsenide and indium arsenide antimonide semiconductor region
If the 2029 date holds, practical quantum computing arrives ahead of most existing forecasts.
Potential risks and opportunities
Risks
- Microsoft's 2029 public commitment creates enterprise planning cycles; a timeline slip would force organizations that built quantum-ready roadmaps around it to rework architectures and vendor strategies
- Other quantum hardware vendors face implicit pressure to make similarly specific timeline claims in response, increasing the risk of competitive overpromising across the sector
- Agentic AI co-design of critical hardware introduces new verification and audit challenges if Discovery's design contributions to Majorana 2 cannot be fully traced or explained after the fact
Opportunities
- Materials suppliers and fabs capable of working with lead-based superconductors and indium arsenide antimonide compounds stand to benefit if Majorana 2's architecture scales toward volume production
- Quantum software developers now have a concrete 2029 utility horizon from Microsoft, giving companies building on quantum platforms a firmer investment thesis for Series A and B rounds
- Microsoft's Discovery agentic AI gains a high-profile hardware design credential with Majorana 2, potentially unlocking enterprise interest in AI-assisted R&D workflows well beyond the quantum space
What we don't know yet
- Whether Microsoft's 1,000x reliability claim for Majorana 2 has been independently verified or submitted for peer review as of the Build 2026 announcement
- What specific problem categories or workloads Microsoft considers within scope for the 2029 'relevant, practical quantum computing' target
- What decisions Microsoft's Discovery agentic AI system made in designing Majorana 2, and how its contributions were validated against human engineering judgment
Originally reported by Tom's Hardware
Read the original article →Original headline: Microsoft Unveils Majorana 2 Quantum Chip at Build 2026 — Designed With Agentic AI, 1,000× More Reliable Qubits, 2029 Utility Target