This year is designated by the United Nations as the International Year of Quantum Science and Technology. This means that the United Nations officially recognizes the transformative potential of quantum science and technology in developing sustainable solutions in the fields of energy, education, communication, and human health, considering it closely related to achieving sustainable development goals. Although cutting-edge research in quantum computing is still flourishing in the laboratory, a global industrial ecosystem centered around quantum computing is gradually taking shape. Anchoring the goal of 'truly practical', quantum computing is accelerating its exit from the laboratory and into real life. Now may be the best time to enter the field of quantum computing. Ten years ago, people were still skeptical whether quantum computing was just a 'trial' in the laboratory, but now it is forming a global ecosystem, "said Laurent Prost, co-founder of French quantum startup Alice Bob. Krista Svore, Chief Research Manager of Microsoft Quantum Research Team (QuArC), straightforwardly believes that quantum computers are "already operational". But in reality, global competitors are still exploring different technological paths separately, hoping that their solutions will ultimately succeed, which means that the research path of new quantum computing technologies has not yet entered the convergence stage. The most important question is - what can quantum computers be used for? So far, its practical use has been quite limited. To fulfill its promise of "solving problems that traditional computers cannot handle," quantum computers not only need to be able to run complex calculations, but also have a sufficiently low error rate to ensure meaningful computational results. But these two goals are mutually constraining: increasing the quantum bits (the basic units of quantum computers) to improve computing power, usually also means a skyrocketing risk of errors. The evolution of qubits from "can be made" to "made well". Scientists attempt to combine multiple "physical qubits" into more powerful "logical qubits" in order to achieve real-time detection and repair of errors during the computing process. You must be able to calculate and correct errors at the same time, "said Svore, which has become a common goal of major research teams: to create as many logical qubits as possible. From the current perspective, American startup atomic computing is temporarily leading in the number of bits achieved. Their latest developed quantum computer has a quantum bit count of 1180, composed of ultracold neutral ytterbium atoms. The French company Pasqal followed suit, adding 1110 atomic qubits (not yet put into computation). Researchers from the University of Science and Technology of China have demonstrated relevant achievements in using artificial intelligence to accelerate atomic assembly. Quantum computing has made rapid progress, and we have moved from the stage of 'can we make it' to the stage of 'can we make it better'. ”Ben Bloom, founder and CEO of Atomic Computing, believes that the neutral atom route is currently in a leading position. However, the breakthrough in quantum computing is far from being as simple as stacking bits. The key to making a truly useful quantum computer is to build a comprehensive system, "said Nicholas Harrigan, Nvidia Quantum Product Marketing Manager. Although Nvidia has not independently developed quantum chips, it is collaborating with multiple companies to study how to better leverage the performance of quantum computing. Other traditional computing industry giants also have similar ideas, for example, Microsoft collaborated with atomic computing last year to launch a commercial quantum machine with 24 logical qubits, which is seen as the first step towards practical quantum devices. But in the competition of logical bits, there is another winner. QuEra, an American quantum computing startup, has showcased over 40 logic qubits. The top scorer is Quantinuum, a well-known quantum computing company in the United States, which successfully ran 50 logical bits. Its president and CEO, Rajib Hazra, revealed that the company's upcoming next-generation quantum computer will have encoding capabilities one trillion times higher than existing records. The competition is intense, with multiple paths showcasing their strengths. Quantinuum company adopts the "ion trap" route, which uses charged ytterbium ions constrained by electromagnetic fields to construct quantum bits. This technological path has also been favored by British quantum startup Oxford Ionics and American quantum computing company IonQ. John Gamble, Senior Director of System Architecture and Performance at IonQ, pointed out that the common advantage of neutral atoms and ion traps is that the connections between quantum bits are more flexible and easier to faithfully execute various algorithms, including converting physical bits into logical bits to achieve different ways of error correction - now the focus is on flexibility and versatility. Because of this flexibility, companies that choose neutral atom and ion trap routes believe they have the opportunity to surpass Google and IBM in the future. Google was one of the earliest companies in the field to claim to have achieved 'quantum superiority'. This concept was first proposed by theoretical physicist John Pryskill at the California Institute of Technology in 2012, referring to the ability of quantum computers to achieve things that traditional computers cannot. In 2019, Google announced the creation of the first quantum computer with computing power far exceeding that of traditional supercomputers. Although this claim was questioned at the time, Google once again claimed quantum superiority in 2024, with its quantum chip "Willow" capable of completing a computing task that would take a traditional supercomputer ten years (1025) to complete in just five minutes. Google and IBM have identified superconducting quantum chips as their main research focus. This solution has the advantages of fast running speed and higher reliability in some scenarios, but it also has limitations. For example, neutral atomic bits are prone to "deviation" from laser controlled states, leading to errors. However, the biggest problem with superconducting technology may be "connectivity". Superconducting quantum bits can usually only be connected to adjacent bits, which makes it difficult for many new error correction algorithms to implement and limits the exploration space. New error correction codes are constantly emerging, and now is far from the end. ”Gamble said. Bloom also stated that the reason why atomic computing companies have shifted from other paths to neutral atoms is because neutral atoms appear to have more advantages in addressing the most core challenges of quantum computing. The superconducting route, once considered the most promising, may be facing a bottleneck. Of course, this does not mean that Google's efforts are worthless. Google's exploration has shown that by integrating more physical bits into logical bits, it can indeed improve error correction capabilities - a key step towards achieving large-scale quantum computers. Currently, IBM's Condor chips have 1121 superconducting qubits, which is only 59 fewer than the highest record set by atomic computing companies. However, IBM's plan is to break through the 4000 bit mark by 2026. To achieve this goal, IBM is developing modules that connect existing chips to create a larger scale "modular" quantum computing platform, and hopes to use this to execute more complex error correction algorithms. American quantum integrated circuit developer Rigetti Computing has not given up on the superconducting route. David Rivas, the Chief Technology Officer of the company, pointed out that superconducting quantum computers have certain practical value. The company has not only launched a buy and use 9-qubit quantum computer, but also provides access to an 84 qubit large-scale processor. These devices are currently being sold to government laboratories and commercial clients. Alice Bob Company also uses superconducting routes to construct quantum bits, but their design philosophy is different: they hope to significantly reduce the error rate of physical bits before constructing logical bits. The company's researchers believe that this approach can achieve completely error free quantum computing with only a few thousand qubits, while competitors may require millions. At present, they have not demonstrated any logical qubits, but their goal is to build a truly practical quantum computer by 2030. What is the endpoint of becoming a low-level technology in society? In the field of quantum computing, the "Five Year Plan" seems to have become a convention. American startup PsiQuantum has a more aggressive plan: to skip the small-scale bit experiment phase and directly launch a large-scale, supercomputer like quantum computer by 2027. They use photons as quantum bits and focus on integrating traditionally complex control components such as lasers and lenses onto industrially mass-produced semiconductor chips. The co-founder and Chief Scientific Officer of the company, Pete Shabolt, stated that they excel at creating difficult but reasonably feasible timelines. Other companies that take the photon route are relatively stable. In 2021, Canadian quantum computing company Xanadu showcased a photon quantum computing chip that can run multiple algorithms. French company Quandela has launched a 12 bit quantum computer with a modular design that facilitates future expansion and upgrades. Who has the most hope of standing out? John Pesky, a professor at the California Institute of Technology who has long been interested in the industry, tends to bet on neutral atoms. He believes that this technology has the ability to scale up and flexibly connect, and has great potential in executing quantum algorithms. "If tens of thousands of neutral atomic qubits can be produced, its performance can be comparable to hundreds of thousands of superconducting qubits. However, the best quantum bit may be the one that 'no one even notices its existence'. Perhaps the most ideal future is not when a certain technology dominates, but when no one cares about the underlying technology anymore. Just as today's AI developers won't be fixated on which CPU to use, future engineers won't need to worry about which physical qubits to use. ”Bloom said that it was not until then that quantum computers truly began to solve problems that could change the world. Although not all quantum experiments are successful, quantum computing is accelerating out of the laboratory and heading towards the real world. (New Society)