A few days ago, the University of Science and Technology of China announced that a research team composed of Pan Jianwei and Lu Chaoyang of the University of Science and Technology of China, in cooperation with the Shanghai Institute of Microsystems of the Chinese Academy of Sciences and the National Center for Parallel Computer Engineering and Technology, has constructed a quantum computing prototype with 76 photons and 100 modes. The "Nine Chapters" of the machine has realized the rapid solution of the "Gaussian Bose sampling" task with practical prospects.
According to media reports, the quantum computing system can process Gaussian Bose sampling a trillion times faster than the current fastest supercomputer; its speed is faster than the 53 superconducting bit quantum computing prototype "Plalanus" released by Google last year Ten billion times. Related papers were published online in the international academic journal "Science" on December 4.
Although there are many terms such as "quantum hegemony" and "one trillion times" in media reports, the current quantum computer is more of a scientific research nature and does not have the ability to fully replace classical computers. It will only be classical computing. Supplement, not substitute. There is a long way to go to build a standard quantum computer that can completely replace classical computers.
Why is quantum computing favored by many people?
In recent years, traditional computer development has gradually encountered a series of problems such as power wall and communication wall, and Moore's law has begun to fail, the growth of chip performance has begun to slow down, and the performance of traditional computers has increased Both the potential and the growth rate are much lower than before. In this situation, the demand for high-performance computing technology that explores new physical principles has emerged. The basic principle of
quantum computing is to use qubits as the basic unit of information encoding and storage, and complete computing tasks through the controlled evolution of a large number of qubits. Quantums in an entangled state are generally selected, such as the two polarization states of photons, the two spin states of electrons, and the two energy levels of ions to construct qubits.
Many people must have heard of Schrodinger’s cat, and qubits are similar to the state of that cat. A qubit can be in state and 1 state at the same time. When the state of the quantum system changes, the superimposed states can change. Compared with transistors that can only express sum 1, qubits can express sum 1 at the same time. This gives quantum computers a huge advantage in parallel computing. After all, N qubits can store 2 to the Nth power at the same time. Data, the amount of data increases exponentially with N. At the same time, one quantum computer operation is equivalent to the effect that an electronic computer has to perform 2 Nth operations, which is equivalent to an evolution equivalent to the completion of parallel processing of 2 Nth data, which is not available in current electronic computers. Possessed. In theory, quantum computing has outstanding advantages in fields such as nuclear explosion simulation, code breaking, materials, and micro-nano manufacturing.
"Quantum Domination" standard quantum computer belongs to "forever 50 years"
In recent years, various news about quantum computers have been heard constantly. The team of Academician Pan Jianwei of the University of Science and Technology of China and the team of Academician Guo Guangcan, overseas Google, IBM, and Intel have successively developed A prototype of quantum computing.
Previously, Google, Intel, and IBM had launched prototypes with 49, 50, and 72 physical bits respectively. Prior to this, the "universal quantum computer" Bristlecone previously released by Google had reached the world's first number of qubits. However, the so-called "universal quantum computer" is actually called a "universal quantum computer". In fact, it is a special-purpose computer that cannot be universal. The "calculation" result of
"Nine Chapters" is also simulated by Monte Carlo algorithm, not a standard quantum computer. Current quantum computers have a subversive effect on specific algorithms and can solve specific problems, but they are not universal and can be used as a supplement to classical computers, but they cannot replace classical computers.
IBM's IBMQ prototype is hailed as "the world's first integrated general-purpose approximate quantum computing system designed for scientific and commercial use, and is currently the world's only independent commercial quantum computer." In some media reports, comparing IBMQ with ENIAC, they think it will usher in a new era of quantum computing.
However, the implementation possibilities of standard quantum computers are similar to those of controlled nuclear fusion, which are both "forever 50 years". Quantum computers will soon replace classic electronic computers but it is just an unreliable fantasy.
What is "forever 50 years"?
If a calculation is very difficult to realize, it is almost impossible to realize it with the current technical reserves, but if it is realized enough to trigger a technological revolution, then it is talking aboutHow long will it take to implement this technology is often expressed in 50 years-after 10 years, the same reporter asks the same expert again, and the response is still 50 years.
also formed the so-called "forever 50 years".
If the answer is 20 or 30 years, but the time is up, but not done, because the specific person in charge is still on duty or alive, it will bring a series of negative factors, but if it is 50 years, it will not be the same as the project that advocated this technology. The leader is no longer there, and the long time will help dilute people's memory. The
quantum computer is such a technology. If it can be realized, it has a broad prospect. However, in terms of the current technical reserves, a standard quantum computer can be produced. Also belonging to the "forever 50 years" is the technology of controllable nuclear fusion. If human beings can succeed in getting rid of the shackles of fossil energy, it may also open up a communist social model. The problem is that this technology belongs to "forever 50 years."
is extremely difficult to implement a standard quantum computer
is extremely difficult to implement a standard quantum computer, mainly because there are huge gaps in many aspects such as coding, system expansion, logic gate accuracy, and coherent cancellation. The
coding is the biggest obstacle to making general-purpose quantum computers. If you want to make a standard quantum computer that exceeds the classical computer, you must obtain at least 50 logical bits. However, the quantum error correction mechanism known to mankind must use 80-10,000 physical bits to generate one logical bit. At present, no commercial company or scientific research institution in the world can make a logical bit. What
wants to show is that 50 logical bits are not the end, but just a new starting point. Although some quantum computer advocates claim that 50 qubits can subvert classical computers, in fact, this statement is not objective.
Marcos Lehrer, a physical chemist at the Royal Institute of Technology Zurich in Switzerland, said: "If we have more than 200 logical bits, we can do things that traditional computers cannot do in quantum chemistry. If we have 5000 logical bits, Bits, quantum computers will bring subversive changes in this field." Z1z
More logical bits need to expand the system, but the difficulty of system expansion will increase by geometric multiples, not to mention that even a logical bit cannot be realized now. The accuracy of
logic gates is also a big problem. To manufacture a standard quantum computer, both single-bit gate accuracy and double-bit gate accuracy must meet very high requirements. The single-bit gate accuracy and double-bit gate accuracy of some of the so-called "quantum computers" actually do not meet the requirements of standard quantum computers.
Quantum computing relies on quantum entanglement. As the environment and time change, this entanglement effect will be lost between quanta. The stable time for maintaining quantum entanglement is really useful, and IBM can only maintain 90 milliseconds before. Such a short time is basically of no practical value.
These so-called "quantum computers" are often called "universal" in publicity, but they can only accomplish specific tasks.
For example, a certain domestic quantum computing prototype can only do Bose sampling, which is not universal. Foreign Google, IBM, and Intel machines are similar. They are all dedicated machines. In some specific applications, they can gain advantages over classic computers, but they are not universal.
For quantum computers, whether it is a commercial company or the government, it is very good to invest a certain amount of money to explore. On the road of technological progress, continuous achievements are also commendable.
However, some commercial companies have overhyped the concept of quantum computing, and ostentatiously exaggerated some limited and partial progress to mislead the public. It seems that quantum computers are about to turn on the eve of revolution, as if classic computers are about to become electronic waste.
Standard quantum computing has a long way to go.
The efforts of Google, IBM, Intel and Academician Pan, and Academician Guo are all commendable.
However, the problem is that the concept of quantum computer has been overhyped. Quantum has become the same as nano, Alot, 5G and other vocabulary, which has been hyped by the media and capital. It has been praised too high, has exceeded its actual technical level, and has derived many Concepts that were not originally available. This approach is not conducive to the healthy development of the industry and technological progress.
due to the development of standard quantumComputers are very difficult. Some people in the industry believe that the so-called standard quantum computer of
is also the definition of a group of people. It may not necessarily be the final implementation method. We must take a more open attitude to see if this system can be implemented Beyond non-quantum computing power.
Coincidentally, IBM engineers also put forward the concept of "approximate quantum computing", that is, when coding problems cannot be solved, they develop algorithms that can adapt to and tolerate noise and get the correct answer.
This is like in a general election. After statisticians ignore some wrong electronic ballots, they still get the correct election results.
In short, quantum computers will inevitably first appear in the form of quantum dedicated machines and gain a foothold in specific fields. These dedicated machines will only be a supplement to the classic computer, not a revolution to the classic computer. It is a very difficult task if
is to make a standard quantum computer that is versatile and has performance beyond classical computers.
