Microsoft unveiled Majorana 2 on June 2, its next-generation topological quantum chip featuring a lead-based superconductor that delivers qubits 1,000 times more reliable than those in its predecessor. Mean qubit lifetime reaches 20 seconds — with some instances persisting up to one minute — compared to the microsecond lifetimes typical of competing approaches; Microsoft technical fellow Chetan Nayak put it simply: “We’re 1,000 times better.” The chip also supports one-microsecond operations and packs each qubit into 1/100th of a millimeter, a size advantage that contributes to scalability. The key materials shift — from the aluminum superconductor used in Majorana 1 to lead, which shields fragile qubits from cosmic disturbances — is credited with the bulk of the reliability gains. With these milestones, Microsoft has revised its forecast: it now expects to deliver a commercially valuable, scalable quantum computer by 2029, cutting its original timeline in half.
Underpinning Majorana 2’s development is Microsoft Discovery, the company’s AI-driven Frontier R&D platform, which also reached general availability on June 2. The quantum team used autonomous AI agents throughout the chip’s development — automating qubit measurements that previously took weeks each, synthesizing nearly two decades of cross-disciplinary data to surface correlations no single researcher could identify, and detecting a miscalibrated temperature sensor that had been corrupting fabrication results. “Agentic AI has permeated almost everything we do,” Nayak said. Microsoft simultaneously launched a Microsoft Discovery App in early preview — downloadable for free and runnable locally with a GitHub Copilot account — extending the platform’s core capabilities to individual researchers outside enterprise settings.