To hear some tell it, quantum computing progress will soon stall, ushering in a "quantum winter" when big companies ice their development programs and investors stop lavishing investments on startups.
"Winter is coming," Sabine Hossenfelder, a physicist and author working for the Munich Center for Mathematical Philosophy, said in a November video. "This bubble of inflated promises will eventually burst. It's just a matter of time."
There are signs she's right. In 2022, quantum computing hit a rough patch, with share prices plunging for the three publicly traded companies specializing in the potentially revolutionary technology. Startups seeking strength in numbers are banding together, a consolidation trend with eight mergers so far by the reckoning of Global Quantum Intelligence analysts.
But you'd have been hard pressed to find a whiff of pessimism at Q2B, a December conference about the business of quantum computing. Industry players showed continued progress toward practical quantum computers, Ph.D.-equipped researchers from big business discussed their work, and one study showed declining worries about a research and investment freeze.
"I don't think there will be a quantum winter, but some people will get frostbite," Global Quantum Intelligence analyst Doug Finke said at Q2B.
Quantum computing relies on the weird rules of atomic-scale physics to perform calculations out of reach of conventional computers like those that power today's phones, laptops and supercomputers. Large-scale, powerful quantum computers remain years away.
But progress is encouraging, because it's getting harder to squeeze more performance out of conventional computers. Even though quantum computers can't do most computing jobs, they hold strong potential for changing our lives, enabling better batteries, speeding up financial calculations, making aircraft more efficient, discovering new drugs and accelerating AI.
Quantum computing executives and researchers are acutely aware of the risks of a quantum winter. They saw what happened with artificial intelligence, a field that spent decades on the sidelines before today's explosion of activity. In Q2B interviews, several said they're working to avoid AI's early problems being overhyped.
"Everyone talks about the AI winter," said Alex Keesling, CEO of quantum computer maker QuEra. "What did we learn? People are trying to adjust their messaging...so that we avoid something like the AI winter with inflated expectations."
Kicking the quantum computing tires
Those quantum computing applications emerged over and over at Q2B, a conference organized by quantum computing software and services company QC Ware. Although quantum computers can handle only simple test versions of those examples so far, big companies like JP Morgan Chase, Ford Motor Co., Airbus, BMW, Novo Nordisk, Hyundai and BP are investing in R&D teams and proof-of-concept projects to pave the way.
The corporate efforts typically are paired with hardware and software efforts from startups and big companies like IBM, Google, Amazon, Microsoft and Intel with big bets on quantum computing. Underpinning the work is government funding for quantum computing research in the US, France, Germany, China, Australia and other countries.
While conventional computers perform operations on bits that represent either one or zero, quantum computers' fundamental data-processing element, called the qubit, is very different. Qubits can record combinations of zeros and ones through a concept called superposition. And thanks to a phenomenon called entanglement, they can be linked together to accommodate vastly more computing states than classical bits can store at once.
The problem with today's quantum computers is the limited number of qubits -- 433 in IBM's latest Osprey quantum computer -- and their flakiness. Qubits are easily disturbed, spoiling calculations and therefore limiting the number of possible operations. On the most stable quantum computers, there's still a better than one in 1,000 chance a single operation will produce the wrong results, an error rate that's disgracefully high compared with conventional computers. Quantum computing calculations typically are run over and over many times to obtain a statistically useful result.
Today's machines are members of the NISQ era: noisy intermediate-scale quantum computers. It's still not clear whether such machines will ever be good enough for work beyond tests and prototyping.
But all quantum computer makers are headed toward a rosier "fault-tolerant" era in which qubits are better stabilized and ganged together into long-lived "logical" qubits that fix errors to persist longer. That's when the true quantum computing benefits arrive, likely five or more years from now.
Quantum computing hype
Quantum computing faces plenty of challenges on the way to maturity. One of them is hype.
Google's captured attention with its "quantum supremacy" announcement in 2019, in which its machine outpaced conventional computers on an academic task that didn't actually accomplish useful work. John Preskill, a Caltech physicist who's long championed quantum computing, has warned repeatedly about hype. Nowadays, companies are focused on a more pragmatic "quantum advantage" goal of beating a conventional computer on a real-world computing challenge.
The technology could be big and disruptive, and that piqued the interest of investors. Over the past 14 months, three quantum computer makers took their companies to the public markets, taking the faster SPAC, or special purpose acquisition company, route rather than a traditional initial public offering.
First was IonQ in October 2021, followed by Rigetti Computing in March and D-Wave Systems on August.
The markets have been unkind to technology companies in recent months, though. IonQ is trading at half its debut price, and D-Wave has dropped about three quarters. Rigetti, trading at about a 10th of its initial price, is losing its founding CEO on Thursday.
Although quantum computer startups haven't failed, some mergers indicate that prospects are rosier if teams band together. Among others, Honeywell Quantum Solutions merged with Cambridge Quantum to form Quantinuum in 2021; Pasqal merged with Qu&Co in 2022; and ColdQuanta -- newly renamed Infleqtion -- acquired Super.tech.
Quantum computing reality
But the reality is that quantum computing hype isn't generally rampant. Over and over at Q2B, quantum computing advocates showed themselves to be measured in their predictions and guarded about promising imminent breakthroughs. Comments that quantum computing will be "bigger than fire" are the exception, not the rule.
Instead, advocates prefer to point to a reasonable track record of steady progress. Quantum computer makers have gradually increased the scale of quantum computers, improved its software and decreased the qubit-perturbing noise that derails calculations. The race to build a quantum computer is balanced against patience and technology road maps that stretch years into the future.
For example, Google achieved its first error correction milestone in 2022, expects its next in 2025 or so, then has two more milestones on its road map before it plans to deliver a truly powerful quantum computer in 2029. Other roadmaps from companies like Quantinuum and IBM are equally detailed.
And new quantum computing efforts keep cropping up. Cloud computing powerhouse Amazon, which started its Braket service with access to others' quantum computers, is now at work on its own machines too. At Q2B, the Novo Nordisk Foundation -- with funding from its Novo Nordisk pharmaceutical company -- announced a plan to fund a quantum computer for biosciences at the University of Copenhagen's Niels Bohr Institute in Denmark.
It's a long-term plan with an expectation that it'll be able to solve life sciences problems in 2035, said physicist Peter Krogstrup Jeppesen, who left a quantum computing research position at Microsoft to lead the effort.
"They really, really play the long game," said Cathal Mahon, scientific leader at the Novo Nordisk Foundation.
What could cause a quantum winter?
Some startups are seeing the frosty investment climate. Raising money today is more challenging, said Asif Sinay, chief executive of Qedma, whose error suppression technology is designed to help squeeze more power out of quantum computers. But he's more sanguine about the situation since he's not looking for investors right now.
Keeping up with technology roadmaps is critical for startups, said Duncan Stewart of the Business Development Bank of Canada, which has invested in quantum computing startups. One of them, Nord Quantique in Quebec, "will live or die based on whether they meet their technical milestones 18 months from now," he said.
But startup difficulties wouldn't cause a quantum winter, Quantinuum Chief Operating Officer Tony Uttley believes. Two scenarios that could trigger a winter, though, are if a big quantum computing company stopped its investments or if progress across the industry stalled, he said.
The quantum computing industry isn't putting all its eggs in one basket. Various designs include trapped ions, superconducting circuits, neutral atoms, electrons on semiconductors and photonic qubits.
"We are not close to a general purpose quantum computer that can perform commercially relevant problems," said Oskar Painter, a physicist leading Amazon Web Services' quantum hardware work. But even as a self-described cynical physicist, he said, "I'm very convinced we're going to get there. I do see the path to doing it."