Summary
- Microsoft has unveiled its Majorana 2 quantum chip, boasting a reliability increase of 1,000 times over its predecessor, aiming for scalable quantum computing by 2029.
- AI technologies were pivotal in expediting research and enhancing materials and manufacturing processes for the chip.
- This announcement heightens concerns regarding the potential of quantum computers to undermine current cryptographic standards.
Artificial intelligence is playing a crucial role in hastening advancements in quantum computing, as researchers endeavor to develop machines capable of challenging the cryptographic systems that protect Bitcoin and much of the internet.
At its annual Build conference on Tuesday, Microsoft presented the Majorana 2, a new topological quantum chip that the company claims is 1,000 times more reliable than its earlier version. The new chip reportedly achieves average qubit lifetimes of 20 seconds, with some qubits lasting up to one minute.
In a blog post following the announcement, Microsoft stated that its Microsoft Discovery platform and agentic AI tools were instrumental in analyzing years of quantum research, discovering promising materials, automating measurements, optimizing fabrication processes, and identifying manufacturing flaws that enhanced qubit reliability.
“By leveraging recent advancements in agentic AI tailored to expedite the scientific process and enhance collaboration, Microsoft's quantum team is overcoming critical challenges related to reliability, speed, and size that have previously hindered the practical application of quantum computing,” Microsoft noted.
Majorana 2 improves upon Microsoft's Majorana 1 chip by substituting its aluminum-based topological superconductor with a lead-based design, which provides better protection for qubits against interference. This modification has resulted in significant gains in both reliability and speed, bringing scalable quantum computing closer to realization—now expected by 2029.
“We need to achieve annual improvements that will lead us closer to delivering a computer that we believe will provide immense commercial and societal benefits,” stated Microsoft Technical Fellow Chetan Nayak. “We must continue progressing along that roadmap, but relative to last year, we are 1,000 times better.”
To assist researchers from various countries and disciplines in navigating the expanding knowledge base of the project, Microsoft's quantum team has developed an AI agent that organizes, analyzes, and highlights information throughout the program.
“Utilizing agentic AI for automating measurements was transformative,” remarked Zulfi Alam, Microsoft's corporate vice president for quantum. “It processes calculations and identifies the optimal conditions for functionality, making simultaneous voltage adjustments that would be impossible for a human to perform due to our linear thought processes.”
This announcement arrives amid ongoing worries about "Q-Day," the moment when a quantum computer may possess the capability to break widely used public-key cryptography, enabling attackers to derive private keys from public ones and potentially steal assets.
Bitcoin is anticipated to be among the most significant targets in such an event, with estimates suggesting that approximately $461 billion worth of BTC could be at risk due to exposed public keys. Developers are actively seeking solutions to address this looming threat.
“What a quantum computer could achieve, particularly concerning Bitcoin, is the ability to forge the digital signatures currently utilized,” Justin Thaler, research partner at Andreessen Horowitz and associate professor at Georgetown University, has previously explained to Decrypt. “A quantum computer could authorize transactions draining your Bitcoin accounts without your consent, which is the primary concern.”
Microsoft is not alone in reporting rapid advancements in this field. In October, Google's Willow chip showed significant reductions in quantum error rates, while recent studies from Caltech indicated that breaking elliptic-curve cryptography might require fewer quantum resources than previously thought.
Google has projected that Q-Day could occur by 2032, while other researchers have suggested it could happen as early as 2030.