Redmond, Washington – Microsoft unveiled its latest topological quantum processor, dubbed Majorana 2, on Tuesday. The chip’s qubits show a mean coherence time of roughly 20 seconds, a thousand times longer than the previous generation. The announcement follows a development sprint that leveraged Microsoft’s internal „Discovery” agentic‑AI platform. Microsoft now projects a fully scalable quantum computer by 2029, cutting its earlier roadmap in half.
The new chip builds on the company’s earlier Majorana prototype, which struggled with qubit decoherence measured in microseconds. By integrating topological error‑correction designs with AI‑driven optimization, engineers extended qubit lifetimes dramatically. Microsoft claims the AI system identified material configurations and control pulse sequences that human designers had missed. The result is a device that can maintain quantum information for tens of seconds, a timescale previously reserved for only the most advanced laboratory setups.
Majorana 2’s 20‑second coherence window translates to a reliability increase of roughly 1,000‑fold over its predecessor. „Achieving this level of stability is a milestone for the entire quantum community,” said a Microsoft spokesperson. The longer coherence time reduces the need for frequent error correction, easing the hardware overhead required for practical algorithms. Industry analysts note that such stability could accelerate the transition from experimental demonstrations to real‑world applications.
Microsoft attributes the rapid progress to its Discovery platform, an agentic AI that autonomously proposes and tests design variations. The system reportedly evaluated thousands of configurations in weeks, a task that would have taken months of manual work. Critics caution that AI‑driven research still depends on human validation, but the company’s results suggest a promising synergy. If the approach scales, it could reshape how quantum devices are engineered across the sector.
The implications extend beyond Microsoft’s own roadmap. A quantum computer with reliable, long‑lived qubits could unlock more complex simulations in chemistry, materials science, and cryptography. By aiming for a scalable system by 2029, Microsoft hopes to position itself as a leader in a market that many predict will be worth billions within a decade. Continued advances in AI‑assisted design may further compress development cycles, bringing practical quantum advantage closer to reality.
What makes topological qubits different from other quantum bits? Topological qubits store information in the global properties of a material, making them inherently more resistant to local disturbances than conventional superconducting qubits.
How does agentic AI differ from traditional AI in chip design? Agentic AI operates autonomously, generating hypotheses, running simulations, and iterating designs without direct human prompting, whereas traditional AI typically assists human engineers with specific tasks.
Will Microsoft’s 2029 timeline guarantee a usable quantum computer? The 2029 target refers to a scalable architecture, not a fully commercial product. Achieving practical quantum advantage will still depend on software development and integration with existing computing ecosystems.