Microsoft says new quantum chip 1,000 times more reliable than predecessor

Microsoft claims its new quantum chip is 1,000 times more reliable than its predecessor, with qubits surviving for an average of 20 seconds instead of milliseconds [1]. The company predicts a quantum machine capable of solving commercially viable problems within three years [1]. The chip, named Majorana 2, uses lead instead of aluminum as a superconductor, improving its effectiveness [1]. Microsoft's approach is based on 'topological' qubits, which it argues are more stable [1]. The company says it will have a quantum machine by 2029 that can solve commercially viable problems, according to corporate vice president Zulfi Alam [1]. However, such a device would require millions of qubits, while the current chip has only 12 [1]. Microsoft's focus on topological qubits has been controversial. The company was forced to retract a 2018 paper in the journal Nature claiming evidence for the Majorana quasi-particle [1]. Despite skepticism from some experts, executive vice president Jason Zander stated, "We stand behind it 100%" [1]. The company is part of a final-stage quantum development program run by the US defense research agency Darpa, with which it has shared all its data for assessment [1]. The global race to develop quantum computing is driven by its potential to tackle problems beyond the reach of classical computers [1]. Microsoft's claimed timeline suggests its quantum computers could eventually address complex issues like removing microplastics or developing better fertilizers [1]. However, qubits remain extremely fragile, and keeping them stable is a fundamental industry challenge [1]. While Microsoft is advancing its specialized hardware, other tech giants are pursuing different computational frontiers. Google DeepMind, a subsidiary of Alphabet Inc., is a leading artificial intelligence research lab known for models like AlphaFold for protein folding and the Gemini family of large language models [5][7]. The Gemini architecture is trained natively on multiple data types to process and generate text, code, images, audio, and video [6]. This highlights the diverse technological pathways—from quantum hardware to advanced AI software—being explored to push computational boundaries.