first practical – After decades of promise, quantum computing is moving from milestones and demos toward early, specific uses—from predicting material behavior to simulating particle physics processes and speeding parts of AI workflows. Even so, experts warn the path to systems
The first time quantum computing seems real, it’s rarely a grand breakthrough. It’s a smaller moment: a calculation runs faster than anyone expects. or a processor models something classical computers struggle to simulate. or a lab team finds coherence holding just long enough to make a result worth chasing.
That’s the mood Peter Zoller, a pioneering theoretical physicist, brings to the effort to build industrial-scale quantum machines. He compares today’s race to become the first to conquer the hardest “peak” to the 20th-century obsession with Mount Everest—exciting in the climb. sharper still when you finally reach the top. “When you’re climbing, you look around worrying, ‘Who is number one?’” he says. “When you reach the top, that’s when you ask yourself, ‘Why the hell did we actually do this?’”.
What comes after “the top” is now beginning to take shape: not another abstract proof, but the first practical targets quantum computers could eventually earn.
The idea traces back to 1995. when Zoller and Ignacio Cirac. then a postdoctoral researcher in Zoller’s group at the University of Colorado Boulder. proposed the first realistic blueprints for a quantum computer. Their approach used trapped ions as qubits—the quantum analogue of digital bits. capable of existing in superposition. representing 0. 1. and states in between at the same time. Long before that. physicists Paul Benioff and Richard Feynman had independently suggested that machines harnessing quantum weirdness could. in theory. outperform classical computers on some tasks.
Today. teams across the world are building larger quantum processors using different physical qubit platforms: ions. neutral atoms. superconducting loops. and more. IBM and Atom Computing. a Berkeley. Calif.–based company. are described as leading the charge with quantum computers hosting more than 1. 000 qubits. And last year. a research group at the California Institute of Technology reported building a record-breaking array of more than 6. 000 qubits.
One reason the timeline has started to feel different is that hardware is no longer stuck at a handful of qubits. “It’s an exciting time because people are fielding quantum computers with hundreds and thousands of qubits. ” says John Martinis. a Nobel Prize–winning quantum physicist. professor emeritus at the University of California. Santa Barbara. and co-founder of quantum hardware company Qolab.
In 2019. Google researchers led by Martinis reported their 53-qubit processor. Sycamore. had become the first to achieve “quantum advantage. ” completing a calculation in 200 seconds that they estimated would take the best classical supercomputers around 10. 000 years. That classical comparison was later disputed: IBM argued its best classical computer could perform the task in just two and half days. Even if the original estimate held up. the calculation was still described as having only academic interest as a proof of principle.
Kihwan Kim. a quantum physicist now at the Institute for Basic Science in South Korea. puts it bluntly about what comes next. “Google’s 2019 demonstration of quantum [advantage] was an important milestone. but many people would say that it did not yet constitute a breakthrough on a problem of broad practical significance.”.
Yet the gap between “proof” and “usefulness” is shrinking in pockets. “We can now simulate things like superconductivity, artificial photosynthesis and small drug designs,” says Michelle Simmons of Silicon Quantum Computing.
Experts say the leap to genuinely useful problems that classical computers can’t reach will likely require qubit numbers to jump significantly—potentially to a million or more—and for qubits to stay coherent for longer. Error correction also becomes essential: qubits must maintain their fragile quantum properties while calculations run. and errors introduced during computation must be addressed.
Even so, near-term progress is already showing up in real tasks. In March. a multi-institutional group reported IBM’s superconducting Heron processors could accurately predict the results of neutron-scattering experiments measuring the structure of a specific antiferromagnetic crystal to an unprecedented scale. using 50 qubits or fewer. The physicists emphasized a constraint: classical computers could perform the same feat faster and more accurately.
So the question isn’t whether quantum computers are doing anything at all. It’s whether they can do it well enough—and reliably enough—that their special capabilities matter.
That tension runs through the biggest promise people still attach to the field: cracking encryption.
Cryptography is the headline-grabber because it ties quantum computing to everyday systems—bank transfers, cryptocurrencies, and digital communication. The core claim is familiar: quantum computers may someday break RSA encryption, a long-standing protocol. Long thought to require processors with at least a million qubits. that estimate has recently been challenged from a surprising direction.
In February. a team from Iceberg Quantum in Sydney. Australia. calculated that with careful optimization and error correction. hackers might need fewer than 100. 000 qubits to break encryption. In March, Google announced a new commitment to migrating its systems by 2029 to protect them from quantum hacking.
Iceberg’s claims have not yet been peer-reviewed, but they’ve still drawn attention. Artur Ekert, a cryptography expert at the University of Oxford, says the argument is credible and has stirred debate. “Many think that the threat to encryption from quantum computers is just mumbo jumbo—and I have also been skeptical—but it could just take a few more papers like this one to make the notion of breaking RSA relevant. ” he says. Ekert also points out his own skepticism has been tested.
Martinis’s own view has shifted too. “If you are worried about RSA encryption—as you should be—I would say it might be broken in five to 10 years,” he says.
The mechanism behind RSA is described as straightforward in one direction and punishing in the other: it’s easy to create a secret key by multiplying two large prime numbers. but it’s effectively impossible for a classical computer to determine the key by efficiently factoring the product back into those primes. Ekert explains that while classical systems can test successive numbers sequentially. remembering each value and looking for a pattern. this approach becomes infeasible with large numbers. Modern classical algorithms use other methods but remain inefficient because execution time increases exponentially with the size of the number to be factored.
Quantum computers aren’t constrained the same way. Qubits can take multiple values simultaneously and become entangled, amplifying their computational power. Ekert says, “Essentially it can cover all possible computational paths at the same time.”
In 1994, theoretical computer scientist Peter Shor, now at the Massachusetts Institute of Technology, proposed that a hypothetical quantum computer could crack RSA encryption using the algorithm Shor developed.
At the same time, work on quantum-resistant cryptographic algorithms is already moving ahead. In 2024, the U.S. National Institute of Standards and Technology (NIST) published three such schemes. Zoller argues this kind of effort largely defuses the threat because it suggests the world can move away from RSA before quantum hacking arrives. “Shor’s algorithm may ultimately be remembered as a landmark scientific achievement of great historical importance for inspiring the development of quantum computers. as much as for its implications for breaking encryption. ” he says.
Ekert is less reassured. He points to a cautionary example from last year: a computer scientist in China proposed—a proposal Ekert says turned out incorrectly—a quantum algorithm capable of breaking NIST’s top candidate. called lattice-based encryption. He notes it took “the brainpower of the whole quantum cryptography community [more than a week] to find a mistake.” His message is simple: the race can turn on whether the next attempt is correct. “Maybe next time it will be correct.”.
If encryption is still a future fear, quantum machines are already finding success in a different kind of work—modeling particle interactions that sit at the heart of fundamental physics.
Michelle Simmons describes the motivation in a way that sounds more like engineering than speculation: “It goes back to Feynman, who articulated that you can’t really understand how nature works unless you build it at the same length scale.”
Daniel González-Cuadra. a quantum physicist at the Institute for Theoretical Physics in Austria. explains why classical computers hit a wall. Simulating interactions of multiple particles rapidly becomes impossible because the information needed to describe these systems grows exponentially with system size. Capturing that complexity requires an equally complex quantum machine.
One benchmark behind the progress is coherence—how long qubits maintain superposition. In 2021, Kim’s China-based team demonstrated trapped-ion qubits can maintain coherence for more than an hour, which he says was “a very important benchmark for scaling up meaningful quantum simulations.”
A clearer taste of quantum simulation is arriving in real-time modeling. In the past year, two teams independently published real-time simulations of matter and antimatter creation during a process called string breaking.
According to the Standard Model of particle physics. González-Cuadra and others describe strongly interacting subatomic particles. such as quarks. as behaving like pairs connected by an elastic string. Pedram Roushan of Google Quantum AI in Santa Barbara uses a vivid image: “like a violin string that vibrates.” His team ran the simulation on Google’s Sycamore chip. which uses superconducting loops as qubits. The simulation tracked how pulling two particles apart increases string tension until it snaps. releasing stored energy by generating new pairs of matter and antimatter particles.
Roushan says, “These theoretical concepts were known since the 1970s, but we were able to visualize them and take a picture of the strings and their breaking.”
The Sycamore experiment is described as a digital simulation, performed on a multipurpose chip with circuits of qubits designed to do many different tasks.
González-Cuadra, Zoller, and colleagues took a different route. Working with a team at QuEra Computing in Boston. they created an analog simulator: a lattice of neutral-atom qubits specially built to simulate string breaking. González-Cuadra says these two string-breaking simulations are among the first to model particle interactions in two spatial dimensions. “The physics is richer, so we could see how these strings fluctuate,” he explains.
Sergio Boixo, Google Quantum AI, points to the direction of travel. “We’re optimistic that we’re going to see the first practical applications in five years.”
These simulations aren’t meant to replace particle physics experiments. Instead, they’re positioned as tools to help physicists hone theories, produce testable predictions, and check them at particle accelerators. So far, they have simulated only simple models that can also be checked with classical computers.
But González-Cuadra argues that quantum simulators could begin to surpass classical counterparts within a couple of years, marking true quantum advantage in practice rather than just on paper.
That promise comes with a serious question: how do you verify that quantum simulations are correct?. Last year. Zoller and his colleagues posted a preprint on arXiv describing a strategy for developing an analog quantum machine that not only makes predictions but also quantifies uncertainty in those predictions. “If you ask me what the big challenge for quantum simulation is, it is the frontier of verification,” Zoller says.
Simulation, though, is only part of the story. Zoller imagines a shift from passive “discovery mode” to an active “design mode,” where quantum computers generate recipes for synthesizing new molecular structures with desirable properties.
That is where the stakes begin to look economic as well as scientific. Simmons frames it as a tool for drugs and batteries. including the prospect of avoiding rare earth elements that can be costly and environmentally damaging. Martinis ties the incentive to how expensive materials science really is. “These things take billions of dollars. so if you just make something even a few percent cheaper or a few percent better. then that’s really worth it. ” he says. “It could be a huge thing not just monetarily but for changing how things are built to make them more ecological.”.
Room-temperature superconductivity sits at the center of many dreams. Superconductivity means electricity flows without resistance, but it typically requires cooling to extremely low temperatures. Some researchers hope quantum engineering can help find new superconducting materials that don’t need cooling at all.
Henrik Dreyer of Quantinuum in Munich says the scale is already daunting for classical modeling: “These systems consist of 1023 particles. whereas classically we can model only about 100 particles.” He says quantum processors would need error rates reduced to just one in a million to make room-temperature engineering feasible. At the moment, he explains, the best chips are down to slightly below one in 1,000.
Dreyer and colleagues have been performing digital simulations of cuprate superconductors using Quantinuum’s Helios chip, which uses 98 trapped-ion qubits. In carefully controlled lab conditions. firing these materials with a laser can briefly create a superconducting state at a relatively high temperature. “The first question is: Why?” Dreyer says.
Last year. Quantinuum posted an arXiv preprint reporting that a two-dimensional simulation of the material shows that under laser fire. its electrons pair up—an essential condition for superconductivity. Dreyer puts the next hurdle in stark terms: “The ultimate question is. ” he asks. “Can we engineer it to do this at room temperature for a minute. an hour. 10 days. or more?”.
Simmons and her team are also building other analog simulation platforms. In Australia. SQC has developed a simulation system called Quantum Twins: a 2D array of 15. 000 clusters of phosphorus atoms embedded in silicon. In February, the team reported that the platform can simulate the transition between insulating behavior and metallic conduction.
Simmons says the capability opens doors to a broader set of targets. “We can now start to simulate things like superconductivity, different battery materials, artificial photosynthesis and small drug designs.”
Google’s quantum efforts are described as spanning multiple industries. Boixo says Google Quantum AI has collaborated with BASF on battery design, Sandia National Laboratories in Albuquerque on fusion energy, and Covestro, a German chemical company, on pharmaceutical development.
Last year. Google implemented an algorithm for modeling molecular structure on Willow. Google’s 105-qubit superconducting processor. which can be used in combination with nuclear magnetic spectroscopy. Boixo says the technique runs 13,000 times faster on Willow than an equivalent algorithm would run on the best classical supercomputer. The method bounces signals onto qubits and listens for their echoes. a design that allows results to be corroborated by another quantum machine.
Boixo calls it “Quantum Echoes,” and says it’s “the world’s first quantum-verifiable algorithm with quantum advantage.” He adds, “We’re optimistic that we’re going to see the first practical applications in five years.”
Then there is the last frontier that’s both promising and controversial: using quantum processors to help classical artificial intelligence.
Jacob Biamonte, an expert in quantum machine learning at ÉTS Montreal, jokes about hype when quantum and AI are paired. But the field is still moving. Simmons says last year SQC launched Watermelon, a quantum-enhanced AI processor to speed up machine learning.
The background matters because classical AI is already strong at finding patterns in huge datasets—useful for optimizing communications and energy networks. Simmons says Watermelon builds on classical reservoir computing. a method that takes input data points and transforms them onto a higher-dimensional neural network to make pattern-finding easier.
SQC’s quantum technique draws on a 2017 prediction by scientists in Japan that classical nodes of a neural network could be replaced by a smaller number of qubits subject to quantum interference. Simmons says the advantage of a quantum reservoir is an exponential increase in dimensionality. enabling a quantum reservoir to achieve the same training results as a classical reservoir but potentially faster and with fewer resources.
Watermelon’s first commercial trial is described as a collaboration with Australian telecommunications company Telstra. Telstra already uses AI to monitor latency and bandwidth patterns. Training models with standard classical methods takes about three weeks. With Watermelon’s help, Telstra achieved the same training results in just two days. Simmons says that matters because data centers are power hungry at scale. and similar optimizations could be rolled out to other energy-intensive tasks. including training AI for image recognition. fraud detection and market prediction. She adds, “I feel like I’m in this freight train that’s going at superhigh speeds.”.
Ekert remains cautious about long-term benefits. “Turning classical data into a quantum form is terribly inefficient,” he says. He argues that the most useful blends of quantum computing and machine learning are already where the work is practical: physicists using classical AI to design quantum error-correcting codes and better quantum hardware.
Last year, for example, Finnish company QMill launched a classical AI service for compressing quantum circuits, reducing the number of gates needed for operation by 20 to 50 percent.
Biamonte thinks the current vision is too small if the end goal is using quantum computers to do machine learning for classical data. He says. “If the goal is to use quantum computers to do machine learning for classical data. it doesn’t even make sense. because classical machine learning is already so good.”.
His alternative is where the excitement is different: if quantum processors one day analyze quantum data directly. “There should be these wonderful patterns that classical computers cannot detect because there are just too many data for their memory,” Biamonte says.
He imagines a scenario where quantum AI could riff on the molecular structure of an existing patented drug to generate multiple configurations with the same benefits. then assess which molecules could be synthesized and patented before a company committed funds to trying. “That’s the exciting future that doesn’t exist yet,” he says.
At the heart of all these threads—simulation, material discovery, encryption timelines, and AI speedups—is the same idea Zoller returns to with his Everest analogy: the moment you finally reach the top, the real question becomes what you’re building it for.
Zoller’s “killer apps” may not arrive as a single, dramatic device that replaces everything else. Instead. they’re showing up as narrow advantages—sometimes in physics. sometimes in materials. sometimes in computation that reduces time and energy use—proof that quantum is not just climbing toward possibility. It’s beginning to earn its place.
quantum computing qubits quantum advantage IBM Heron Google Sycamore encryption RSA NIST quantum-resistant cryptography quantum simulation string breaking superconductivity quantum AI Watermelon reservoir computing