Major breakthrough: IBM overcome the “unreliable” problem in quantum computing
Major breakthrough: IBM overcome the “unreliable” problem in quantum computing
Major breakthrough: IBM overcome the “unreliable” problem in quantum computing.
on June 14th, four years ago, Google claimed to have achieved “quantum supremacy”, which aroused people’s attention to quantum computing.
However, in practical applications, quantum computing always has a “reliability” problem.
The same calculation is repeated many times, and it is very likely that it will come up with a different answer each time. On Wednesday, “Big Blue” IBM claimed that they have found a way to solve the reliability of quantum computing.
It should be noted that the basic binary units of traditional computers are 0 and 1, either 0 or 1.
But in quantum computing, its basic unit of measurement, the qubit, can be both 0 and 1, and it can also be both 0 and 1.
This phenomenon is called quantum superposition. Quantum computers realize the function of storing a large amount of information at the same time through quantum superposition.
As a result, they can quickly store large amounts of data, explore multiple possibilities, and choose the most efficient solution when dealing with complex tasks.
However, since maintaining the superposition state of qubits is very difficult, the slightest environmental change (vibration, electric field, magnetic field, cosmic radiation) can also cause the superposition state to collapse, causing calculation errors. Therefore, the world has not yet been able to create a quantum computer that is error-free and has a wide range of uses.
On June 14, 2003, IBM researchers announced that they have devised a way to manage the unreliability of quantum computing to produce reliable, useful answers.
IBM scientists have published a research paper in the journal Nature, titled “Evidence for the Practicability of Quantum Computing Before Fault Tolerance.” Error-tolerant quantum computing refers to quantum computing protected by quantum error correction.
Papers published by IBM
In 2019, researchers at Google claimed that they had achieved “quantum supremacy,” which means that quantum computing has computing power that exceeds that of all classical computers.
However, IBM criticized Google at the time, believing that Google exaggerated the performance of quantum computing and misled the public.
On Wednesday, IBM researchers said they had found some new and more useful methods, albeit with lower-key names.
“We’re entering a phase of quantum computing that I call ‘practical,'” said Jay Gambetta, vice president of IBM’s quantum business. “The era of practicality.”
Dorit Aharonov, a professor of computer science at the Hebrew University of Jerusalem who was not involved in the research, said: “What IBM is showing here is really progress towards serious quantum algorithm design. It’s an important step in the direction of the movement, which is surprising.”
How to reduce the error?
In the new study, IBM researchers performed a different task that has piqued the interest of physicists. They used a quantum processor with 127 qubits to simulate the behavior of a 127-atom-scale bar magnet in a magnetic field. These rod magnets are small enough to be governed by the peculiar rules of quantum mechanics. This is a simple system known as the Ising model, which is often used to study the ferromagnetism of matter.
IBM’s quantum processor used in experiments
The problem is too complex to compute an exact answer on even the biggest and fastest supercomputers. But on a quantum computer, the calculation can be done in less than a thousandth of a second. However, each quantum calculation is unreliable, because the fluctuation of quantum noise (referring to the fluctuations that exist in any monochromatic light) will inevitably interfere with the calculation and cause errors, but each calculation is very fast, so Can be repeated.
In fact, in many calculations, researchers deliberately added extra noise, making the answers even less reliable. But by varying the amount of noise, the researchers could infer the specific characteristics of the noise and its impact at each computational step.
“We can amplify the noise very precisely, and then we can rerun the same circuit,” says Abhinav Kandala, IBM’s manager of quantum capabilities and demonstrations and one of the authors of the Nature paper. We get these results for different levels of noise, and we can extrapolate the results without the noise.”
Essentially, the researchers were able to remove the effects of noise from unreliable quantum computing, a process they call “error mitigation.” “You have to get around the noise by inventing really clever ways to mitigate its effect,” Dr. Aharonov said. “That’s exactly what they did.”
How accurate is it?
To arrive at the answer for the overall magnetization produced by the 127 magnet bars, IBM’s quantum computer performed a total of 600,000 calculations. How accurate is the answer?
For help, the IBM team turned to physicists at the University of California, Berkeley. Although the Ising model with 127 magnet bars is too large, with too many possible configurations, to fit into a conventional computer, classical computer algorithms can produce approximate answers. This technique is similar to JPEG image compression when less important data is discarded to reduce file size while preserving most of the image’s details.
IBM Quantum Computing Researchers
Michael Zaletel, a physics professor at UC Berkeley and one of the authors of the Nature paper, said that when he started working with IBM, he thought his classical computer algorithms would do better than quantum ones. better. “The results were a little bit different than I expected,” Dr Zaleter said.
The results show that quantum computers can accurately solve certain configurations of the Ising model. On simpler examples, the classical and quantum algorithms agree. For more complex but solvable instances, quantum and classical algorithms produce different answers, but the quantum algorithm gives the correct answer.
IBM Quantum Experimental Research Laboratory
Sajant Anand, a graduate student at the University of California, Berkeley, has done a lot of work on classical approximation research. Based on the above experimental results, he believes that for other situations where the results of quantum computing and classical computing are inconsistent and the exact solution is not known , “There is reason to believe that the results of quantum computing are more accurate”.
It is unclear whether quantum computing can undisputedly outperform classical computing in the Ising model. Anand is currently trying to add a reduced-error version of the classical algorithm that has the potential to match or exceed the performance of quantum computing.
“There’s no clear indication that they’ve achieved quantum supremacy here,” Zaleter said.
Temporary solution
In the longer term, quantum scientists expect a different approach, error correction, to detect and correct computing errors, which will open the door to many uses for quantum computers.
Currently, error correction methods are already being used to fix errors in conventional computers and data transmission. But error correction may still be years away for quantum computers, requiring better processors to handle more qubits.
The IBM scientists believe that error mitigation is a temporary solution that can now be used to solve increasingly complex problems outside the Ising model.
“It’s one of the simplest physical science problems out there,” Dr. Gambetta said, “so it’s a great start. But the question is, how do you generalize it to solve more interesting physical science problems?” “These problems could include figuring out the properties of exotic materials, accelerating drug discovery and simulating fusion reactions.
