Research advances toward quantum supremacy with impossible results in classical computing technology


For science, the future of quantum computing is already here, it’s a reality even with current limitations. Investigation published nature This Wednesday shows that a 127-qubit processor, a capacity already found in current commercial quantum computers, is capable of measuring the values ​​expected in physical processes, even with the errors that this type of calculation still generates. For Göran Wenden and Jonas Bilander, researchers at Chalmers University of Technology (Sweden), the work “shows that quantum processors can be useful for some specific computations, despite the errors,” a power that is still a few years away from being realized. This demonstration does not establish quantum supremacy, understood as the ability to solve problems that conventional computers cannot solve, but progresses toward it by demonstrating that one advantage of current quantum computing, the one that is already there, is effective and “exceeds the capabilities of the best current classical computational methods.” with post-analysis failure mitigation processes.

Today’s supercomputers, such as IBM’s Summit, are capable of processing more than 200 billion calculations per second. But a quantum one can go into the trillions thanks to superposition, a property that allows particles to be in two states or any superposition of them. In this way, while two bits (a bit is the minimum unit in classical computing) can store a number, two qubits can contain four and ten bits, and 1024 concurrent states, so the capacity is greatly expanded for each added qubit.

The problem is that quantum superpositions decay until they become classical states (decoherence) by any interaction with the environment: temperature, electromagnetism, vibrations… Any interference produces noise, reduces to microseconds the time that superpositions double computing power and generate failures trying to Mitigate them with programming, looking for particles close to the elusive Majorana that maintain coherence, or avoiding them with very complex systems, isolated and at temperatures of absolute zero (-273°C).

See also  Europe revives balloons to transport platforms into the stratosphere | technology

In this way, fault-tolerant quantum computing is currently impossible for current technologies and is “out of the reach of current processors,” according to the paper’s authors, although companies like Google have recently taken giant steps in this regard.

The demo, now published by IBM researchers Youngseok Kim, Andrew Eddins, and Abhinav Kadala, along with other authors, shows that a quantum processor and post-processor can generate, manipulate, and measure quantum states that are so complex that their properties cannot be exact. Estimated using classic approximations.

It’s not about computing speed, it’s about capacity. “No conventional computer has enough memory to encode probabilities calculated by 127 qubits,” the authors say. Winden and Bilander agree: “The key quantum advantage here is scale rather than speed: encoding 127 qubits is a problem for which no conventional computer has enough memory.”

quantitative advantage

“These experimental results are made possible by advances in superconducting processor coherence and calibration, as well as the ability to characterize and process noise in a controlled manner. These experiments show the usefulness of quantum computing in an era of fault tolerance.” [la actual] and demonstrate an essential tool for implementing short-term quantum applications, the authors say.

The physicometry has already been demonstrated by an international European team in which Professor from the University of Seville Adán Cabello took part, who managed to monitor the quantum state of the strontium ion throughout the process, not just at the beginning and the end. This first film was considered something unprecedented and lasted for a millionth of a second Physicist As one of the most prominent developments of 2020.

See also  Find out what social networks know about you | your technology | Country
One of IBM's first commercial quantum computer prototypes.
One of IBM’s first commercial quantum computer prototypes.ibm

For the current research, the Ising model was used, which is a proposed model for studying particle magnetic transition. But the goal was not the physical process, but to prove that a reliable measurement can be made on a complex system with a commercially available quantum computer, even if it is not fault-tolerant. The question asked by the article nature en: Can we do something useful with quantum computers today, with a small number of qubits and relatively high error probabilities? The authors’ answer is yes, but it has a trick: error mitigation, says Carlos Sabin, a Ramón y Cajal researcher in the Department of Theoretical Physics at the Autonomous University of Madrid (UAM). Science Media Center (SMC).

The authors show that their devices [de IBM]After mitigating the errors, Sabin explains, it provides reliable results when calculating the physical volumes of a system. He adds, “If these results are confirmed (for example, by the competing Google team) it would mean a first step in proving the usefulness of today’s relatively small and noisy quantum computers, when they are helped by error mitigation.” Although this particular calculation has no direct practical application (since the parameter values ​​showing quantum supremacy probably do not correspond to real physical systems), at least, the Ising model [el utilizado en el experimento] It has physical inspiration, so it is possible that there are models of similar complexity with immediate implementations that can also be attacked by similar machines and an approach based on mitigation, not debugging,” he concludes.

Highlights Juan José Garcia Ripoll, Research Scientist at IFF-CSIC Institute for Fundamental Physics smc that it This work, of excellent quality, demonstrates the computational power of IBM’s 127-qubit quantum computer.” García-Ripoll summarizes the conclusions of the research: “Our quantum computers, despite their inaccuracy, can simulate very complex problems of interest to physics . Although the quantum computer makes errors, the protocol allows to cancel them out and obtain very accurate quantum predictions; And simulation techniques in classical computers for problems of this kind produce results that are less accurate than a quantum computer.”

For the Spanish physicist, the result is “not necessarily final”, although quantum computing has developed processors such as the Osprey, also from IBM and with 413 qubits. “It is possible that other scientists will improve the state of the art in tension networks [los sistemas clásicos usados para problemas como el abordado en Nature] And it manages to match or exceed what this processor can do with 127 qubits,” he adds.

A similar view is made by Göran Wendin and Jonas Bylander: “Does this breakthrough improve the prospects for applying quantum computing to problems relevant to industry? The answer: probably not. Such algorithms must involve much more qubits and many more operations to be competitive with classical supercomputers, and such quantum computations will inevitably get drowned out in noise.”

You can write to [email protected] and follow up country technology in Facebook And Twitter Or sign up here to receive The weekly newsletter.



Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button