Last year, Google’s prototype quantum computer completed a mathematical calculation that required 100 years of operations on a state-of-the-art supercomputer within a few minutes, and won wide international praise. This achievement is called "quantum hegemony" (also known as "quantum superiority") by the scientific community-that is, quantum computers have achieved traditional computers that will never reach the level of computing.
Google Quantum Computer
On Thursday, a quantum research team led by Chinese scientist Pan Jianwei published a new record of "quantum hegemony" in the journal Science. It is said that this quantum computer named "Nine Chapters" completed the result that the world's third largest supercomputer took 2 billion years to reach within a few minutes.
Pan Jianwei is manipulating the "Nine Chapters"
. The two quantum computing systems operate in different ways. Google uses ultra-low temperature superconducting metals to build quantum circuits, and a research team at the University of Science and Technology of China in Hefei has achieved experimental results by manipulating photons (particles of light).
Quantum computers are far from ready for actual work. However, there are signs that there are two completely different forms of quantum computing that can surpass supercomputers, which will boost the R&D enthusiasm and investment in this budding industry. Lu Chaoyang, a professor of physics at the University of Science and Technology of China who participated in the project, said that this milestone is a "necessary step" towards "large-scale fault-tolerant quantum computers."
Google and its competitors, including IBM, Microsoft, Amazon, Intel and several large start-up companies, have invested heavily in the development of quantum computing hardware in recent years. Google and IBM announced their latest prototypes on the Internet, while the cloud platforms of Microsoft and Amazon each have a large amount of quantum hardware from other companies (including Honeywell).
IBM quantum computers
The potential capabilities of quantum computers come from their basic building blocks called "qubits." Just like the bits of a traditional computer, they can represent the data of 0 and 1; but qubits can also use quantum mechanics to obtain an unusual state called superposition, which contains two possibilities of 0 or 1. . With enough qubits, it is possible to achieve calculation shortcuts that traditional computers cannot take-this advantage will be enhanced as more qubits work together.
Quantum computers have not yet ruled the world, because engineers have not yet been able to make enough qubits work together reliably. The quantum mechanical effects they rely on are very subtle. Google and the Chinese research team were able to complete their "quantum hegemony" because they successfully gathered a relatively large number of qubits.
Google’s experiment used a superconductor chip called "plane tree", which has 54 qubits and is cooled to an absolute 1 degree. One of the qubits cannot work, but the remaining 53 qubits are enough to prove that it performs better than traditional computers on a carefully selected statistical problem. It is not clear how many high-quality qubits a quantum computer needs to do actual work, and expert estimates range from hundreds of millions to millions. The
qubit
Chinese research team used statistical tests to prove its quantum advantage, but its quantum data carrier adopts the form of photons, which are reflected by an optical mirror and transmitted through an optical circuit placed on the experimental platform. Each photon read at the end of this process can be considered roughly equivalent to reading a qubit on the Google quantum processor, so that the result of the calculation can be obtained. The
researchers stated in the paper that 76 photons were measured from the "Nine Chapters" prototype, but only 43 on average. The members ran the mathematical calculations completed by "Nine Chapters" on the "Shenwei Taihu Light" supercomputer. "Shenwei Taihu Light" is the most powerful supercomputer in China and the third fastest supercomputer in the world. But it is far from comparable. Researchers calculated that it would take more than 2 billion years for traditional supercomputers to complete the tasks completed by the "Nine Chapters" in a little over 3 minutes. One difference between
"Nine Chapters" and Google's "Platanus" is that the photon prototype is not easy to reprogram to run different calculations. Its settings are hard-coded into its optical circuit all at once. Christian Weedbrook, CEO and founder of Toronto quantum computing startup Xanadu, is also working on photon quantum metersCount. He said that this result is still worth noting, and it reminds people that there are many feasible ways to achieve quantum computing. "This is a milestone in photonic quantum computing," he said. "Although it is not perfect, it is good for all of us in the world." Academia and industry are developing several different forms of quantum hardware. Qubits based on superconducting circuits are the most prominent, thanks in part to the huge investments of Google and IBM. The quantum computers provided by industrial giants Honeywell and startups including IonQ are called “ion traps”. They are based on single atoms suspended in an electric field and are made of qubits. Purchase through Amazon and Microsoft cloud services.
ion trap
In September of this year, Wadebrook published his first prototype on the Internet, providing early customers with up to 12 qubits. He said that his team can make more flexible devices than the "Nine Chapters" and believes that photonic quantum computers will soon catch up with other forms of computers. Their advantage is that they use the same components used in many telecommunications networks. Professor Lu of the
Chinese research team said that they are also working on larger and more flexible versions of quantum computers. Other researchers have shown that the techniques used in the group's "quantum hegemony" experiment can be used for chemical calculations or to solve problems involving mathematical graphics, which are very important in areas such as transportation and social networks. Proponents of
photonic quantum computing and "ion trap" quantum computing have stated that their technology should be easier to mass produce than the superconducting chips favored by IBM and Google because they do not need to install the equipment in an ultra-low temperature refrigerator. However, no one knows which form of quantum computing will first reach the level of practical application. "We all have strengths and weaknesses," Wadebrook said.
This article is adapted from "Wired"
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-