Quantum computer brings chaos under control in "landmark" discovery

Physicists have discovered a new way to control "chaotic behavior" inside a quantum computer, hailing the finding as a step toward overcoming one of the biggest obstacles to building reliable quantum machines.

The problem scientists are trying to overcome stems partly from quantum computers' extraordinary complexity, in which every qubit is highly sensitive to its surroundings.

Tiny changes in temperature, stray electromagnetic signals, or hardware imperfections can disturb a qubit's quantum state, introducing errors that accumulate and undermine calculations.

Quantum computers face another problem. As qubits interact, information spreads between them and becomes increasingly scrambled, creating chaotic behavior that is extraordinarily difficult to predict and control.

Now, researchers from Rutgers University, IBM and collaborating institutions have shown that a quantum computer can repeatedly check and reset parts of itself while running, bringing this chaotic behavior under control.

The advance moves us toward larger, more reliable quantum computers that can detect and correct errors while operating.

Addressing the errors of existence

In a study published in Nature Physics, the team demonstrated one of the largest experiments in real-time quantum feedback control, a technique that lets a quantum computer take measurements and respond to the results while continuing to operate.

During their experiment, the team deliberately scrambled quantum information and then repeatedly reset parts of the computer to bring the resulting chaos under control.

The physicists discovered that by performing these resets thousands of times, they could stop the scrambling from taking over.

The finding matters because the more qubits a quantum computer has, the more complicated its behavior becomes. Information can spread between interacting qubits, making the system extraordinarily difficult to track, predict and control.

The researchers demonstrated that even a relatively large quantum system could be brought under control through repeated interventions.

Scaling quantum computers will also require identifying and correcting errors while they are working, a capability known as quantum error correction.

This would allow scientists to build a fault-tolerant quantum computer, a machine capable of continuing its calculations correctly even though its individual qubits are constantly making small mistakes.

Rather than allowing errors to accumulate until calculations become unreliable, a fault-tolerant computer would detect and correct them as it operates.

No one has yet built a fully fault-tolerant quantum computer.

“To build a fault-tolerant quantum computer, we have to make these checks and corrections not just once or twice, but a huge number of times during a calculation,” said Jedediah Pixley, a professor in the Department of Physics and Astronomy at Rutgers University and director of the Center for Materials Theory.

“This experiment is a landmark on that road,” said Pixley, who also is a member of the Flatiron Institute, the scientific research division of the Simons Foundation in New York, and one of the senior authors of the study. “It says, ‘OK, we’ve taken another important step to get there.’”

The work also confirmed a prediction about quantum behavior. As the researchers increased how frequently the computer measured and reset its qubits, the system abruptly shifted from chaos to a state they could control, demonstrating how order can emerge from disorder.

What is a qubit, and why is it so difficult to control?

Conventional computers store information in bits, each representing either 0 or 1.

Quantum computers instead use qubits, which can exist in a combination of the two states, known as a superposition.

A qubit is not necessarily a particular particle or material. Different types of quantum computers use different physical systems to store information.

IBM's quantum computers use microscopic superconducting electrical circuits built onto chips. These circuits contain components called Josephson junctions, which exploit quantum effects to create and control the information-storage states.

The chips are cooled to temperatures close to absolute zero, colder than outer space, allowing the circuits to operate as qubits.

At these extremely low temperatures, the circuits can exist in quantum states representing 0, 1 or a superposition of both.

Multiple qubits can also become entangled, meaning their quantum states become linked in ways that cannot be fully described independently.

These properties give quantum computers their potential power, but they also create enormous complexity.

As more qubits interact, information spreads between them and their combined quantum state becomes increasingly difficult to describe.

In principle, the amount of information needed to describe the complete quantum state of a system can grow exponentially as more qubits are added.

This makes large quantum systems extraordinarily difficult to simulate on conventional computers and creates serious challenges for scientists trying to understand and control their behavior.

The physical hardware introduces further difficulties.

Heat, stray electromagnetic signals, imperfect control pulses and tiny flaws in the circuits can disturb qubits and alter the information they hold.

Over time, these errors can accumulate until a calculation can no longer be trusted.

“The challenge was not simply to measure a qubit, but to do it repeatedly while the rest of the processor continued to operate,” said another senior author of the study, Maika Takita, principal research scientist for quantum error correction and dynamic circuits experiments at IBM. “This experiment shows that present-day hardware can coordinate quantum operations, measurements and resets thousands of times across a large system.”

Unlike ordinary computer data, quantum information cannot simply be copied. Measuring a qubit can also change or destroy the quantum state being used.

This presents scientists with a difficult problem. They need to monitor what is happening inside quantum computers, but directly measuring the information involved can disturb the system they are trying to control.

A fault-tolerant computer must therefore be able to detect and correct errors without destroying the quantum information required to complete its calculations.

The latest experiment did not achieve full quantum error correction. Instead, it demonstrated a technique for controlling the scrambling of quantum information, an important capability for developing more reliable quantum systems.

Towards quantum error correction

The researchers conducted their experiment using an IBM Quantum Heron processor containing 156 qubits, selecting a connected chain of up to 100 qubits.

They programmed the system to alternate between two competing tasks.

The first deliberately created chaos.

The scientists made neighboring qubits interact, causing quantum information to spread through the system and become increasingly scrambled.

The process is similar to repeatedly shuffling a deck of cards. With every shuffle, the original order becomes more difficult to recognize.

In a quantum computer, however, the process is considerably more complicated because information can become distributed across multiple interacting and entangled qubits.

As the scrambling continues, the original information becomes increasingly difficult to recover from individual parts of the system.

The second task introduced a means of control.

The computer measured individual qubits and reset them to a known quantum state, represented as zero.

Have you tried turning it off and on again?

Resetting a qubit does not mean turning it off. Instead, the machine returns it to a predictable starting state, removing the quantum information it previously held.

By repeatedly resetting selected qubits, the researchers could counteract the scrambling and gradually push the system toward a simpler, more orderly state.

Crucially, the rest of the processor continued operating throughout the experiment.

The team performed nearly 5,000 operations that linked pairs of qubits, alongside almost 5,000 measurements and resets.

According to the researchers, this was the largest successful demonstration of the approach to date.

The scientists controlled how often the computer performed each task. By changing the balance between scrambling and resetting, they could determine which process dominated the system.

When scrambling occurred more than half the time, chaos won and the system remained chaotic.

But when measurements and resets occurred more frequently, the researchers could guide the qubits toward the ordered state they wanted.

In effect, the physicists demonstrated that repeatedly returning individual qubits to known states could prevent quantum information from becoming hopelessly scrambled across the computer.

READ MORE: "Harvest now, decrypt later is a today problem": Collapsing uncertainty in quantum security

At approximately equal frequencies, even a slight change in which operation occurred more often caused a sudden transformation across the system.

Physicists call such a dramatic change a phase transition, similar to the process that occurs when water freezes into ice.

In this case, the qubits shifted collectively from chaotic behavior into a state the researchers could control.

“Small changes can lead to large and nearly impossible-to-predict consequences, like the classic idea that a butterfly flapping its wings in Africa could eventually contribute to a hurricane hitting Louisiana,” said Justin Wilson, an associate professor of physics at Louisiana State University and a study coauthor.

“But this out-of-control behavior can be brought under control by randomly intervening, and there is a sharply defined rate at which that happens. It wasn’t obvious that this transition between chaos and control would survive in quantum physics, but our results show that it does.”

The findings suggest that even highly complex quantum systems can be controlled through repeated measurements and interventions.

The researchers demonstrated that they could suppress the scrambling of quantum information across a large system. They did not create a machine capable of automatically correcting arbitrary errors during a calculation.

Nevertheless, the ability to perform thousands of measurements and resets while a quantum processor continues operating represents progress toward the techniques needed for fault-tolerant quantum computing.

From a birthday-party bet to a quantum experiment

The project began with a friendly bet at Pixley's wife's birthday party in 2021.

At the gathering in Highland Park, New Jersey, Pixley's longtime collaborator, Sriram Ganeshan, wagered that a puzzling quantum effect had a counterpart in ordinary physics.

The bet led the researchers to develop a new theory about controlling quantum chaos.

A chance conversation with IBM physicist Barbara Jones later helped them turn their theory into a practical experiment.

Pixley eventually lost the wager and paid Ganeshan two scoops of blueberry ice cream when they met at the Aspen Center for Physics in Colorado.

The IBM processor let the researchers test their theory on systems with up to 100 qubits, far beyond the approximately two dozen qubits they could simulate on conventional computers.

This illustrates the immense complexity of quantum systems. As qubits interact and become entangled, the amount of information needed to describe their combined quantum state can grow exponentially.

Even powerful conventional computers can struggle to model these interactions in full.

Despite this complexity, the experimental results closely matched the researchers' theoretical predictions.

“What was striking was that the transition became more clearly defined as we studied larger systems,” said Haining Pan, a Rutgers postdoctoral researcher at the time of the study and a co-first author.

“The agreement between the experiment and the theory showed that this was not an accidental feature of a small circuit. It was collective behavior emerging across the quantum system.”

READ MORE: Google speeds up post-quantum migration to counter "harvest now, decrypt later" attacks

The findings suggest that repeated measurements and resets can control chaotic quantum behavior even as the number of interacting qubits increases.

Researchers still face substantial challenges before they can build quantum computers capable of detecting and correcting errors throughout long, complicated calculations.

But the experiment demonstrates that scientists can repeatedly measure and manipulate parts of an extraordinarily complex quantum system while it continues operating, bringing order to behavior that would otherwise remain chaotic.

The research involved scientists affiliated with Rutgers University, IBM, the City College of New York, the CUNY Graduate Center, Iowa State University, Ames National Laboratory, Pennsylvania State University and Louisiana State University.

Follow Machine on LinkedIn