As quantum computing remains in its infancy, it's still anyone's game-and cloud giants like Microsoft and Amazon.com are racing to stake their claim.
For Microsoft, establishing dominance means overcoming a past controversy that still shadows its quantum program. In 2018, university researchers from the Netherlands published a paper asserting strong evidence for Majorana zero modes, the elusive, particle-like states central to Microsoft's quantum processors.
Their claims sparked swift backlash from the scientific community, which accused researchers of cherry-picking data. Both the paper and an earlier study from 2017 ultimately were retracted.
Microsoft claims no affiliation with the papers, the products of company-sponsored researchers out of Delft University of Technology in the Netherlands. Internal investigations found no evidence of scientific misconduct. Yet, as rivals like International Business Machines and Alphabet's Google pull ahead in the race, Microsoft's quantum effort remains under a persistent cloud.
Today, the company is distancing itself from the Delft papers, confidently asserting there is much more to its quantum initiative. Its latest chips bear no technical resemblance to the setup referenced in the studies, and the company continues to push ahead with new generations of its Majorana architecture.
Two decades in the making, quantum computing is now Microsoft's longest-running research program. The company opened the doors to a new research center in Maryland on Tuesday, bringing its technology closer to independent evaluators from the Defense Advanced Projects Research Agency (DARPA).
That development coincided with Quantum World Congress, one of the industry's largest gatherings and a proven hype generator. The event serves as a stage for companies to pitch progress and reassure investors they're still in the running.
For Microsoft, that reminder came in the form of a keynote address by Jason Zander, the head of its quantum effort. Having spent over 30 years at the company, Zander usually seeks out the next techology," spends about a decade scaling it, and moves on. Quantum has broken that template.
Zander drew a careful distinction between Microsoft's existing systems, which are tailored for research and development, and the commercial-scale machines of the future. "At the end of the day, we're looking for better benchmarks to build better hardware," he told a swarm of attendees Thursday at Quantum World Congress.
Microsoft has partnered with quantum developer Atom Computing to deliver a new system to Denmark next year. Named Magne-the Danish word for "strength"-the project is backed by a $90 million investment from the Novo Nordisk Foundation and Denmark's Export and Investment Fund.
Powered by Microsoft's quantum software, Magne uses laser-manipulated neutral atoms as its qubits, or basic units of information. It sound like science fiction brought to life. Yet for Microsoft, delivery of the system is one piece of a bigger puzzle.
Beyond developing its own quantum programming language, the company provides cloud access to third-party quantum hardware through Azure Quantum. As it does in classical computing through partnerships with Nvidia and Advanced Micro Devices, Microsoft aims to become a quantum ecosystem enabler.
Other cloud giants have been carving out a space in quantum. Amazon Web Services has partnered with QuEra, another developer working with neutral atoms, to host its systems on the Amazon Braket cloud service.
A quantum computer's performance advantage over classical systems depends entirely on its underlying algorithms, which harness small-scale physics to solve problems. Since 2020, Amazon Braket has allowed users to explore algorithms across multiple quantum modalities. The service offers access to seven devices from five developers today.
Michael Brett, who leads Amazon's quantum go-to-market strategy, acknowledges the technology is still in its early stages. Over the past few years, "we've been operating quantum computers at a loss for the science," Brett said.
Amazon's partners are equally realistic. QuEra's chief commercial officer, Yuval Boger, detailed how the company's weekly meetings with DARPA officials keep them on target as they build more capable systems.
In their current state, "quantum computers are toys," Boger said. "I think they're beautiful, sophisticated toys, but they're toys. But that's changing."
The main drawback of today's quantum computers is their fragility. Highly sensitive to environmental noise like temperature and vibration, they are inherently error-prone. While the industry struggles to agree on benchmarks, the overarching objective is to build "fault-tolerant" machines that can detect and fix hardware errors in real time.
QuEra aims to be "first to utility" with an error-corrected machine by 2028. That beats IBM's aggressive timeline to deliver a fault-tolerant supercomputer by a full year.
QuEra CEO Andy Ory tempered that vision with practical realism. "This industry is about to happen," Ory told Barron's. "So what's all this activity? It's preparation. It's research into algorithms. It's finishing the science."
Given the rapid pace of change, long-term financial forecasts vary widely. Boston Consulting Group projects quantum computing will generate up to $850 billion of economic value by 2040. Estimates from McKinsey, which many insiders view as too bullish, pin that number somewhere between $1.3 trillion and $2.7 trillion by 2035.
Pioneering physicist Richard Feynman once described quantum mechanics as "perilous but still correct." Indeed, the dream of quantum is both a future pursuit and an immediate reality-and while pure-play developers strive to get systems to market, established tech companies may find that building the broader ecosystem is their best path forward.
