SAN FRANCISCO (Reuters) - Microsoft on Tuesday unveiled a new quantum computing chip that it redesigned with the help of AI, saying it now believes it will have commercially useful quantum machines by 2029.
The new target date puts Microsoft on track to have quantum computers the same year as rival IBM, which last month said it plans to spend $10 billion on quantum machines. It also spun out a company to make quantum chips for others, with backing from President Donald Trump's administration.
Microsoft had not previously given a target year for the new chip, saying only that it would be a matter of years, not decades.
AI TOOLS DRIVE MATERIALS BREAKTHROUGH
The biggest change to Microsoft's internally made chip versus its predecessor is that it uses an entirely new set of materials. While Google, IBM and many others make quantum chips with superconducting wires made out of aluminum, Microsoft's will be made out of lead, a larger atom.
Microsoft made the switch with the help of AI tools that it developed for use in materials science, and the result was a 1,000-fold improvement in some aspects of Majorana 2's performance, said Jason Zander, an executive vice president at Microsoft who oversees the firm's quantum efforts. The breakthrough, Zander said, was figuring out how to use lead, which is water soluble, on a chip without the lead washing away during the manufacturing process.
"The reason why people don't use it to build chips is it requires an incredibly specialized process to be able to go figure that out. And we figured it out," Zander said. Microsoft's approach to quantum computing relies on quasiparticles known as Majoranas, which had not been proven to exist until Microsoft claimed to have observed them.
SCIENTIFIC CRITICISM OVER CLAIMS
Its claims have kicked off a flurry of criticism among physicists who say Microsoft has not publicly released enough data to verify its claims. The publication Science last year alerted readers that it was investigating the data used, in an earlier Microsoft study from 2020, and some critics of Microsoft's earlier papers say that the problems with its data and protocols still exist in the research released on Tuesday.