Quantum computing timelines draw skepticism despite new US push
The Trump administration and Microsoft have set near-term goals for useful quantum machines, but researchers say practical uses remain unproven.
By Maya Lindqvist · Senior Technology Correspondent
3 min read
The Trump administration is pressing to accelerate quantum computing as companies promise useful machines within a few years. Independent researchers told The Verge that today’s systems remain too small and error-prone to solve commercial or scientific problems that matter outside the lab.
On June 22, President Donald Trump issued an executive order aimed at speeding the US quantum computing industry as it competes with China. The same day, Trump science adviser Michael Kratsios said he wanted a quantum computer capable of scientific discovery by 2028, according to The Verge.
Microsoft is making its own near-term pledge. In June, the company announced a chip called Majorana 2 and said it had advanced its plan for a scalable, practical quantum computer by 2029. Several independent physicists challenged the claim, including Henry Legg of the University of St. Andrews, who told The Verge the announcement was “complete codswallop.”
Why useful quantum computing remains out of reach
Quantum computers use qubits rather than ordinary bits, allowing them to handle probability-based calculations in ways classical computers cannot. Researchers hope they could eventually simulate molecules for drug discovery, battery materials or other scientific problems that conventional supercomputers struggle to model.
The machines are not expected to replace ordinary computers or become consumer devices. Dries Sels, a Boston University physicist, told The Verge that a quantum computer would serve a specific purpose, likely through cloud access to specialized hardware.
Existing quantum machines cannot yet run long, reliable calculations. The Verge reported that qubits lose information and make errors, and those errors pile up as algorithms become more complex.
Researchers have made progress on error correction, which spreads one unit of information across multiple physical qubits to create a more reliable “logical qubit.” Google reported a logical qubit made from 105 physical qubits in 2024, while IBM, Amazon and Quantinuum later reported designs using fewer physical qubits, according to The Verge.
Andrew Houck of Princeton University also reported a superconducting qubit last November that held information three times longer than the previous record holder. Houck told The Verge the advance came from careful changes to the materials and fabrication process.
Microsoft’s Majorana approach faces dispute
Microsoft’s bet centers on Majorana particles, a proposed basis for qubits that the company says could be easier to scale. Legg argues Microsoft has not shown it has created the particles and says the company’s data could be explained by quantum dots instead.
Microsoft quantum lead Chetan Nayak rejected that criticism. He told The Verge the company stands behind its results and roadmap, and said Microsoft’s papers show it is creating and controlling Majorana particles.
The dispute follows earlier scrutiny of Microsoft-linked work. The Verge noted that Nature retracted a 2021 paper by Microsoft-affiliated researchers that had claimed strong evidence for Majorana particles.
Money keeps flowing
Large technology companies and governments continue to spend heavily. IBM said in June it would invest more than $10 billion in quantum computing over five years, and the Trump administration said in May it planned $2 billion for nine quantum companies, including $1 billion for IBM, according to The Verge.
Google and IBM use superconducting circuit qubits, Quantinuum uses barium ions, and QuEra uses rubidium atoms. The Verge reported that experts view neutral atoms, ions and superconducting circuits as leading qubit types.
Researchers disagree on when a useful result will arrive. Eleanor Crane of King’s College London told The Verge she expects a useful scientific simulation by 2028, while Houck said it is likely before 2035. Legg was more skeptical, saying there is no evidence any platform can scale to useful computation within a decade or perhaps two.
Rajibul Islam of the University of Waterloo summed up the uncertainty more broadly. He told The Verge that quantum computing remains an early technology and that he does not know of any guaranteed application yet.
This story draws on original reporting from The Verge.