In a small chip at ETH Zurich, a team led by Yiwen Chu has shown that the tiny mechanical vibrations of microscopic resonators can serve as the working memory of a quantum computer, much as RAM supports a conventional machine's processor. The idea reflects an old human pattern: separating the act of thinking from the act of holding onto what has been thought, and finding that each can be built from a different material. Reported in Science, the demonstration hints at a path toward machines that are more compact and hold their quantum information longer, though the distance between a working ex
Vibrating Quantum Memory: ETH Zurich Team Uses Mechanical Resonators to Store Quantum Information
Tiny mechanical vibrations serve as compact quantum memory
Mark
Why does it matter that the memory is vibrations rather than electromagnetic fields?
Mimi
Size, mostly. Electromagnetic memory can take up a lot of room, which makes scaling up hard. The mechanical resonators are substantially smaller, and each can hold many vibration patterns.
Mark
And the CPU and RAM comparison, is that just a metaphor?
Mimi
It's the actual design idea. A superconducting qubit does the processing, and a separate memory holds information the qubit retrieves, modifies and returns.
Luke
Careful with the longer storage claim, though. The reporting says vibrations can preserve states longer before fading, but gives no storage times and no comparison figures.
Mimi
Fair. It's described as an advantage, not quantified here.
Mark
What did they actually demonstrate?
Mimi
Two tasks, the quantum Fourier transform and period finding. Both require coordinating multiple quantum states precisely, so they're a meaningful test.
Luke
Meaningful, but still a lab chip. The claim that it can in principle do all operations for arbitrary computations is a statement about capability, not a demonstration of a general-purpose computer.
Mark
So what's unknown?
Luke
Whether it scales. Nothing reported shows the precision holding with more components or harder computations.
Mimi
And that's exactly what the team says comes next.
O Pulso
- A quantum computer's memory has grown too bulky to scale, and researchers in Zurich are searching for a way to shrink it without losing precision.
- Chu's team has replaced electromagnetic storage with vibrating mechanical resonators, a shift that unsettles the usual design of quantum hardware.
- A superconducting qubit now acts as the processor, reading, altering and returning quantum states held in motion on the chip.
- Two foundational tasks, the quantum Fourier transform and period finding, have run successfully on the new architecture, signaling it can compute rather than merely store.
- The approach is still confined to an experimental chip, and its fate now depends on whether precision holds as the systems grow larger.
In a small chip at ETH Zurich, a team led by Yiwen Chu has shown that the tiny mechanical vibrations of microscopic resonators can serve as the working memory of a quantum computer, much as RAM supports a conventional machine's processor. The idea reflects an old human pattern: separating the act of thinking from the act of holding onto what has been thought, and finding that each can be built from a different material. Reported in Science, the demonstration hints at a path toward machines that are more compact and hold their quantum information longer, though the distance between a working experiment and a general-purpose quantum computer remains considerable.
Inside a chip small enough to rest on a fingertip, something is vibrating far beyond the range of human hearing. At ETH Zurich, Yiwen Chu and her team have put those vibrations to work as the memory of a quantum computer, and they have reported the result in the journal Science.
The design mirrors the ordinary computer on a desk. A processor does the calculating, while a separate working memory holds what the work needs as it proceeds. Here, a superconducting qubit plays the processor's role, retrieving stored information, modifying it and returning it to memory. What sets the system apart is the medium. Most quantum computers keep their information in electromagnetic form. Chu's team keeps it in mechanical vibration, held in microscopic resonators, each storing a state as a particular pattern of motion.
The promise lies in three qualities. Electromagnetic memory is precise but bulky, which makes large machines hard to build. Mechanical resonators are far smaller, can host many distinct vibration patterns at once, and hold quantum states longer before they fade.
To test the idea, the team ran the quantum Fourier transform and period finding, procedures that require many quantum states to be coordinated with great care. Passing them showed the system could store, control and maintain relationships among several states at once. The authors argue that, in principle, the architecture can carry out the basic operations needed for general-purpose computation.
Yet the work remains a laboratory chip, and its authors say so. The real test lies ahead: scaling to larger systems while keeping the precision seen in the lab. Chu and her colleagues continue to refine the approach, but practical quantum computers for wide scientific and industrial use are still some distance away.
Inside a chip small enough to sit on a fingertip, something is vibrating. The motion is far above anything a human ear could register, and it is not noise. At ETH Zurich, a team led by Yiwen Chu has put those vibrations to work as the working memory of a quantum computer, and they have described the result in the journal Science.
The design borrows its layout from the ordinary computer on a desk. A conventional machine has a central processing unit that does the calculating and a separate working memory, the RAM, that holds information while the work goes on. Chu's system follows the same division of labor. A superconducting qubit serves as the processor and control unit, playing the part of the CPU. A distinct quantum memory holds the information needed mid-calculation, so the processor can retrieve it, modify it and put it back.
What differs is the medium. Most quantum computers today store information electromagnetically. Chu's team stores it in mechanical vibrations. As she put it, in their quantum working memory the information is not held electromagnetically, as is usually the case now, but as mechanical vibrations. The vibrations live in microscopic structures called mechanical resonators, each holding information in a particular pattern of motion. When the qubit needs a stored state, it reads the vibration, performs its operation and returns the altered state to memory.
The case for the change rests on three advantages. Electromagnetic memory is well understood and can be paired with superconducting qubits to give very precise control over quantum states, but it can take up considerable space, which makes it hard to scale experimental devices into the larger machines researchers and industry want. Mechanical resonators are substantially smaller. They can also support many different vibration patterns, which means more information can be held at once. And mechanical vibrations can preserve quantum states longer before the information fades, which matters for calculations that need data to stay available across many processing steps.
The paper is billed as the first experimental demonstration that mechanical resonators can work together with superconducting qubits to carry out quantum computations. To test it, the team ran two foundational techniques: the quantum Fourier transform and period finding. Both demand that a computer coordinate several quantum states with considerable precision. Igor Kladaric, a doctoral student in Chu's group and co-author of the paper, explained that the Fourier transform is a basic procedure required by many quantum algorithms, and that the period-finding algorithm they implemented showed how it can be used.
Passing those tests means the system stored, controlled and kept coherent relationships among multiple quantum states at the same time. The authors present that as evidence the architecture can support genuine quantum computation, not just act as a memory device. They also report that, in principle, the system can carry out all the fundamental operations needed for arbitrary quantum computations, the essential capabilities of a programmable, general-purpose machine.
The result comes from an experimental chip, and the source is plain about the limits. The next challenge is scaling to larger systems with more processing power while keeping the precision shown in the lab. Whether mechanical memory stays effective as components and computational complexity grow will decide its long-term success. Chu and her colleagues are continuing to study how to develop the approach, and practical quantum computers for wide scientific and industrial use remain some distance away.
Citações Notáveis
In our quantum working memory, information is not stored electromagnetically, as is usually the case today, but in the form of mechanical vibrations.— Yiwen Chu, ETH Zurich
The Quantum Fourier Transform is a fundamental procedure for many quantum algorithms, and the period-finding algorithm demonstrated how it can be used.— Igor Kladaric, doctoral student and co-author
Quer a matéria completa?
Leia o original em Science Daily ↗