China is the only country in the world to achieve quantum superiority on two technological routes, the United States only has one superconductor, while other countries do not have one. In this sense, China's quantum computing research is leading the world.
On October 26, 2021, the national media was spreading a major news: China's quantum computer research has once again made a breakthrough, hundreds of billions of times faster than supercomputers (https://mp.weixin.qq.com/s/_HvxJ3D69tPUUza1AuAqew)! Specifically, the two quantum computers of the team of academician Pan Jianwei of the University of Science and Technology of China have been upgraded, and the "Nine Chapters" and "Zu ChongZhi" have become the second. The conclusion is that China has become the only country in the world to reach the milestone of "quantum computing superiority" in two physical systems.
Chapter IX II No. 1
Chapter IX II No. 2
Zu Chong No. 2
Hey, what do these words mean? What is quantum computing? What is the superiority of quantum computing? What is the superiority of quantum computing in two physical systems? What are the benefits of achieving this? Does this mean that China has surpassed the United States? ......
The last question can be answered briefly: Yes, China has surpassed the United States. But for other technical issues, it is difficult to say clearly in two words. In fact, I recently published a popular science book, A Brief Introduction to Quantum Information, which clearly explains all these issues. Welcome to order this book, after reading you will become an expert in quantum information.
"Quantum Information Shorthand"
"Quantum Information Shorthand" ¥69 purchase
I've been so busy lately that I haven't had time to interpret the news in the first place. However, I know that whenever news of quantum information comes out, there will be media interviews. Sure enough, after a while, the reporters of the English edition of the Global Times called, and I want to praise their professionalism (https://mp.weixin.qq.com/s/lkPGtEtryrmy8UdA4sW32w).
Today, I'm going to take the time to explain the basic picture a little bit.
First of all, what is quantum computing? The answer is a new computational principle that gets results for certain problems much faster than traditional computers. A typical example is factorization, i.e
21 = 3 × 7，
15 = 3 × 5
This decomposition decomposes a natural number into the product of prime factors.
Factorization is a traditional problem. What does this mean? When a number is very small, it is of course easy for you to decompose it, and you can decompose 21 regardless of three seven twenty-one. But when a number is large, say, hundreds or thousands of bits, it becomes difficult to decompose it. Because we don't have a particularly ingenious algorithm, the time it takes to decompose it with a traditional computer rises exponentially with the number of digits. For example, it takes 150,000 years to decompose a 300-digit number, and 5 billion years to decompose a 5,000-digit number!
For quantum computers, however, factorization is a problem that can be solved quickly. In 1994, someone proposed a quantum factorization algorithm, which calculates much more slowly with the number of digits. Also by breaking down 300-bit numbers, quantum algorithms reduce time from 150,000 years to less than a second. Break down 5,000 digits and the quantum algorithm will reduce the time from 5 billion years to 2 minutes!
Now, everyone understands the power of quantum computing, right?
However, there are two points that need to be emphasized.
One is that quantum computers only surpass classical computers for certain problems, not all problems. Some problems classical computers have calculated very quickly, such as addition, subtraction, multiplication and division, and quantum computers have no advantage over them. So the prospect of quantum computers is to use with classical computers, not to replace classical computers. It will never completely replace a classical computer, and both will be used in their own suitable scenarios.
What is generally thought of as a quantum computer
Actual quantum computers
Quantum computers are used in conjunction with classical computers
Second, quantum computer hardware capable of decomposing large numbers has not yet been built. The largest number we can currently decompose with a quantum algorithm is
291311 = 523 × 557，
This was achieved in 2017 by Academician Du Jiangfeng and Professor Peng Xinhua of HKUST. 291311 is only a six-digit figure, far from hundreds or thousands.
A 2017 paper by Du Jiangfeng and Peng Xinhua using quantum algorithms to decompose 291311
In fact, we have not yet built any quantum computers with practical value. So the research status of quantum computers is that software comes first and hardware is the bottleneck. At present, the competition between countries is mainly hardware.
The next question is, what does quantum computing superiority mean? It refers to a problem in which a quantum computer surpasses the strongest classical computer.
What is the strongest classical computer available today? It is Japan's supercomputer Fugaku, which can run 4.42 billion billion floating-point operations per second (https://mp.weixin.qq.com/s/ANC3_b-lm_6UQ9yb9telhA). For example, Fugaku's calculation amount of one second is comparable to the speed of two years for everyone in the world to sleeplessly at a rate of one per second. It can be seen that the current classical computer is very powerful, and it is not easy for a quantum computer to surpass a classical computer.
Top 500 supercomputers in June 2021
Therefore, quantum computers must choose some of their own fast algorithms and classical computers can only surpass classical computers with problems with slow algorithms. This is why the definition of the superiority of quantum computing should emphasize "for a certain problem".
Specifically, there have been only two experiments that have previously achieved quantum superiority.
In 2019, Google achieved quantum superiority with a superconducting system, and their experimental device was called "Sycamore" and the problem they handled was called "random circuit sampling."
Google Paper Figure 1 demonstrates the structure of a quantum computer
In 2020, Pan Jianwei and Lu Chaoyang of HKUST achieved quantum superiority with optical systems, their experimental device is called "Nine Chapters", and the problem they deal with is called "Gaussian boson sampling".
Nine chapters of light quantum interference physical diagram
On December 5, 2020, the author and Professor Yuan Zhensheng introduced nine chapters in nine chapters
What do these two questions mean, you can go to my book "A Brief Introduction to Quantum Information". Here we would like to remind everyone that every quantum computer experiment requires some kind of physical system, superconductivity, optics, ion traps, nuclear magnetic resonance, and so on. Just as classical computers can be implemented with many kinds of physical systems, initially an abacus, then a mechanical computer, then an electron tube, then a transistor, and now an integrated circuit, a quantum computer can also be realized with many kinds of physical systems. Now superconductivity and optics are two of the more popular technical routes, and others are also being studied.
So when you see the news of a quantum computer, you can ask two professional questions: First, what mathematical problem does it deal with? Second, what physical system does it use? As soon as the expert hears you ask these two questions, he knows that you are very good at it. And if you ask some inconspicuous questions, such as "will the quantum computer play a game will be stuck", experts will know that you do not understand at all!
In May 2020, Pan Jianwei and Zhu Xiaobo of HKUST published an important result called "Zu Chong Zhi". Like Google's "planewood," it belongs to superconducting quantum computers. I was interviewed by CCTV's "Common Concerns" (https://mp.weixin.qq.com/s/GaAy15LBfSe-4G9PhAQKuQ) and the English edition of the Global Times (https://mp.weixin.qq.com/s/-rFRZ1TvPJUCGILgODNRkw) and I told them the significance of the matter. Zu Chong zhi surpassed sycamore in some indicators, such as its qubit number of 62, more than sycamore's 53. But Zu Chong did not achieve quantum superiority because it was not maneuverable enough. What it actually does is just to show the two-dimensional walk of the quantum, and in principle any computational task can be achieved with this method.
The University of Science and Technology of China successfully developed the 62-bit programmable superconducting quantum computing processor Zu Chongzhi
Schematic of Zu Chong's two-dimensional superconducting qubit chip, with each orange cross representing a qubit
With these backgrounds, it is understandable that the recent breakthrough is that Pan Jianwei, Lu Chaoyang, Zhu Xiaobo and others upgraded the Nine Chapters and Zu Chong to the second. The Nine Chapters II further expanded its advantages over classical computers, while Zu Chong Zhi II achieved quantum superiority. Therefore, we can claim that China is the only country in the world to achieve quantum superiority on two technological routes, and the United States only has superconductivity, while other countries do not have one. In this sense, China's quantum computing research is leading the world.
Specifically, Chapter Nine Ii upgraded the number of photons from the original maximum of 76 to a maximum of 113, resulting in an increase in the advantage over classical computers from one hundred billion times to one hundred million billion billion times. Someone asked me, why not upgrade to 152, or 76 double? The answer is that the ability of a quantum computer does not grow linearly with the number of photons, but exponentially or even hyper exponentially, so it is not easy to add a photon every time.
Paper data graph a graph represents the dimensions of the output state space b plot shows the advantage multiple of the optical quantum computing prototype over supercomputing
There's an interesting story about a hardcore opponent of quantum computing, called Gil Kalai, an Israeli mathematician. He once thought that boson sampling, the problem done in the nine chapters, would never achieve quantum superiority. His formulation is:
"Imagine an alien army, much more powerful than we are, landing on Earth and asking us to show a boson sample of 5 photons or destroy the planet. In this case, we should mobilize all our quantum engineering forces and try to achieve it. But if aliens are asking for boson sampling, say, 10 photons, then our best bet is to try to attack the aliens. ”
You see, he thinks that boson sampling of 10 photons is impossible! And what we actually did was 76, and now it's 113 again! With this dude's flag, you understand what a great achievement this is, right?
Jill Gloria
The progress of Zu Chong No. 2 is to use a new flip-soldering 3D packaging process to solve the problem of large-scale bit integration, and realize the high-density integration of 66 data bits, 110 coupling bits, and 11 reads, and the maximum spatial dimension reaches 10 to the 19th power. The so-called state space dimension is the number of basic states that can be allowed in this system. The spatial dimension of planewood only reaches 10 to the 13th power, while ZuChong-2 reaches 10 to the 19th power, which is 6 orders of magnitude higher than it. So it is also the problem of performing random line sampling, zu chong no 2 is much more difficult than the plane tree.
Under this premise, zu chong no two has an advantage over classical computers ten million times, while the plane tree is only one million times. That is to say, Zu Chong No. 2 achieved a higher multiplier advantage for a more difficult problem, so it was much stronger than the plane tree.
The fidelity of quantum random line sampling varies with the depth of the line and the time it takes the fastest supercomputer, Fugaku, to complete the same task
Finally, looking to the future, what are the uses of these results? The answer is that it is not useful yet, but with them as the basis, the next stage will be useful. We hope to find practical applications in specific fields within, for example, five years, such as quantum machine learning, quantum chemistry, quantum approximation optimization, and so on.
A target for quantum computers
China has achieved quantum computing superiority in both optical and superconducting technology routes, surpassing the United States, but this is far from the end, but the beginning. The wonderful world of quantum information is waiting for everyone to explore. Finally, I recommend my popular science book "Quantum Information Shorthand" to you again, you can order it in the official micro store of HKUST Publishing House or Jingdong and Tmall, thank you for your support.