Quantum computer will be the focus of computer research in the future. Quantum computer uses the quantum mechanical state of particles to represent information, which can realize complex calculations that cannot be carried out by electronic computers at present. In 2000, German and American scientists developed a five-qubit quantum computer for nuclear magnetic resonance, and successfully passed the experimental operation.
American scientists have developed a new technology that can be used to make DNA computers. It can not only limit the activity range of genetic material DNA molecules to the solid surface for operation, but also greatly simplify the steps of solving complex mathematical problems through DNA operation, marking another step for scientists to develop powerful DNA computers.
A research team composed of American and British scientists successfully developed the world's first DNA "engine". This achievement indicates that in the near future, scientists can manufacture electronic circuits of molecular size to replace the current silicon chip circuits, thus making the future computers smaller, lighter and faster.
Canadian scientists have recently developed an optical chip, which is an important achievement in the journey of developing optical computers.
Working in a virtual office
In the future, your office may not be the Fang Gezi unit now, but more like a holographic deck in a science fiction movie. Computer scientists are experimenting with a technology called remote immersion. This technology will enable us to share an office with colleagues far away and feel that they are close by. ]
In the United States, the official name of this project is "National Remote Immersion Initiative", and computer scientists from several universities are jointly developing it. At present, the prototype of the system has come out. The system allows scientists to meet and communicate with colleagues in other States through the display screen installed at right angles on both sides. It feels like facing a colleague through a window. This technology is different from the early video conference. The pictures provided by the video conference are the same size as the real people and scenes, and it is a three-dimensional (3-D) effect.
In the prototype system, each researcher is surrounded by a group of digital cameras, and his every move can be photographed from different angles. They also wear trackers and similar special glasses to watch stereoscopic movies. When the researcher turned his head, his perspective on his colleagues also changed. For example, if he leans forward, his colleagues on the other side will appear closer, even though they may actually be thousands of miles away.
In order to make this system work effectively, a high-functional computer must be able to quickly convert the digital image of the other party into a digital signal that can be transmitted through the Internet, and then restore the signal to an image at the receiving end and display it on the screen. At present, the whole operation process is still not ideal, because the speed, memory and other functions of the computer are still incompatible.
In the near future, with the improvement of computer technology, scientists hope that long-distance compatibility technology can also enter other application fields. For example, with this technology, patients in remote areas can let famous doctors in big cities "see" the disease. Let's wait and see.
Quantum computer is a physical device that follows the laws of quantum mechanics, performs high-speed mathematical and logical operations, and stores and processes quantum information. When a device processes and calculates quantum information and runs quantum algorithms, it is a quantum computer. The concept of quantum computer comes from the research of reversible computer. The purpose of studying reversible computers is to solve the energy consumption problem of computers.
The quantum computer was first proposed by richard feynman, which originated from the simulation of physical phenomena. It can be found that when simulating quantum phenomena, the amount of data becomes huge because of the huge Hilbert space. The operation time required for complete simulation becomes considerable, even unrealistic astronomical figures. At that time, richard feynman thought that if a computer composed of subsystems was used to simulate quantum phenomena, the operation time could be greatly reduced, and the concept of quantum computer was born.
Quantum computers, or, by extension, quantum information science, were mostly in an armchair strategist state in 1980' s. It was not until PeterShor put forward the quantum factorization algorithm in 1994 that quantum computers became a hot topic, because it could crack the RSA encryption algorithm prevailing in banks and networks. In addition to theory, many scholars focus on using various quantum systems to realize quantum computers.
Semiconductors record and calculate information by controlling integrated circuits, while quantum computers hope to control the state of atoms or small molecules and record and calculate information.
In the 1960s and 1970s, it was found that energy consumption would lead to the chip heating in the computer, which greatly affected the integration of the chip, thus limiting the running speed of the computer. It is found that energy consumption comes from irreversible operation in the calculation process. So, does the calculation process have to be completed by irreversible operation? The answer to the question is that all classical computers can find corresponding reversible computers without affecting their computing power. Because every operation in a computer can be converted into reversible operation, it can be expressed by unitary transformation in quantum mechanics. In fact, the early quantum computers were classical computers described in the language of quantum mechanics, and did not take advantage of the essential features of quantum mechanics, such as superposition and coherence of quantum states. In classical computers, the basic information unit is bits, and the operation objects are various bit sequences. Similarly, in a quantum computer, the basic information unit is a quantum bit, and the operation object is a sequence of quantum bits. The difference is that the quantum bit sequence can be not only in the superposition state of various orthogonal States, but also in the entangled state. These special quantum states not only provide the possibility of quantum parallel computation, but also bring many wonderful properties. Different from classical computers, quantum computers can do arbitrary unitary transformation, and after obtaining the output state, they can measure and get the calculation results. Therefore, quantum computing greatly expands the classical computing, which can be regarded as a special kind of quantum computing in mathematical form. The quantum computer transforms each superimposed component, all these transformations are completed at the same time, and they are superimposed according to a certain probability amplitude, and the results are given. This kind of calculation is called quantum parallel calculation. In addition to parallel computing, another important use of quantum computer is analog subsystem, which is beyond the power of classical computer.
1994, PeterShor, an expert from Bell Laboratories, proved that quantum computers can perform logarithmic operations, and the speed is much faster than that of traditional computers. This is because quantum, unlike semiconductors, can only record 0 and 1, and can represent multiple states at the same time. If the semiconductor is compared to a single musical instrument, the quantum computer is like a symphony orchestra, which can handle many different situations in one operation. So a 40-bit quantum computer can solve the problem that 1024-bit computer took decades to solve.
One: The basic concept of quantum computer
[Edit this paragraph]
Quantum computer, as its name implies, is a machine to realize quantum computing. To understand quantum computing, first look at classical computing. A classical computer can be physically described as a machine that transforms the input signal sequence according to a certain algorithm, and the algorithm is realized by the internal logic circuit of the computer. A classic computer has the following characteristics:
Its input state and output state are both classical signals, which are described in the language of quantum mechanics, that is, their input state and output state are both eigenstates of a mechanical quantity. If the binary sequence 011010 is input, the quantum notation is used, that is, | 0110 >. All input states are orthogonal to each other. It is impossible to input the following superposition states into a classical computer: c1| 010/kloc-0 >+C2 |1001>.
Every step of the transformation in the classical computer evolves into an orthogonal state, but the general quantum transformation does not have this property. Therefore, the transformation (or calculation) in the classical computer only corresponds to a special set.
Corresponding to the above two limitations of classical computers, quantum computers are extended respectively. The input of a quantum computer is described by a quantum system with a finite energy level, such as a two-level system (called a qubit), and the transformation of a quantum computer (that is, quantum computation) includes all possible unitary transformations. Therefore, the characteristics of quantum computers are:
The input state and output state of quantum computer are general superposition States, which are usually not orthogonal to each other;
Transformations in quantum computers are all possible unitary transformations. After obtaining the output state, the quantum computer measures the output state and gives the calculation result.
It can be seen that quantum computing greatly expands classical computing and is a special kind of quantum computing. The most essential features of quantum computing are quantum superposition and quantum coherence. The transformation of each superimposed component realized by quantum computer is equivalent to a classical calculation. All these classical calculations are completed at the same time, superimposed according to a certain probability amplitude, and the output results of quantum computer are given. This kind of calculation is called quantum parallel calculation.
Whether it is quantum parallel computing or quantum analog computing, quantum coherence is essentially used. Unfortunately, it is difficult to maintain quantum coherence in practical systems. In a quantum computer, qubits are not an isolated system, they will interact with the external environment, leading to the attenuation of quantum coherence, that is, decoherence (also known as decoherence). Therefore, to make quantum computing a reality, a core problem is to overcome decoherence. Quantum coding is the most effective method to overcome decoherence. The main quantum coding schemes are: quantum error correction code, quantum error avoidance code and quantum error prevention code. Quantum error-correcting code is an analogy of classical error-correcting code, and it is the most studied code at present. Its advantage is wide application range, but its disadvantage is low efficiency.
So far, there is no real quantum computer in the world. However, many laboratories around the world are pursuing this dream with great enthusiasm. There are many schemes to realize quantum computing, but the problem is that it is really too difficult to manipulate micro quantum States in experiments. At present, the proposed scheme mainly uses the interaction between atoms and optical cavities, cold trap binding ions, electron or nuclear spin vibration, quantum dot manipulation, superconducting quantum interference and so on. It is hard to say which scheme is more promising, but quantum dot scheme and superconducting Josephson junction scheme are more suitable for integration and miniaturization. In the future, the existing schemes may be useless, and finally a brand-new design will emerge, which is based on a new material, just like semiconductor materials are for electronic computers. The purpose of studying quantum computer is not to replace the existing computer with it. Quantum computer makes the concept of computing completely new, which is the difference between quantum computer and other computers such as optical computer and biological computer. The function of quantum computer is far more than solving some problems that classical computers can't solve.
Quantum computers perform a series of large-scale and high-precision operations through quantum splitting and quantum patching. Its floating-point operation performance is incomparable to the CPU of ordinary home computers. The large-scale operation mode of quantum computer is actually similar to the batch program of ordinary computer. Simply put, its operation mode is high-speed quantum repair through a large number of quantum splits, but its accuracy and speed are beyond the reach of ordinary computers, so the cost is quite amazing. At present, the only quantum computer is still in Microsoft's Silicon Valley hometown, still in the experimental stage, and it will take some time to put it into use. Of course, quantum computer doesn't let us play video games, because it is like cutting paper with a laser cutter. Its main uses are, for example, measuring the precise coordinates of stars, quickly calculating the volume of irregular three-dimensional graphics, accurately controlling robots or artificial intelligence and other tasks that require large-scale, high-precision and high-speed floating-point operations. Behind this series of difficult operations are terrible energy consumption, short service life and terrible heat.
Assuming that 1 ton of uranium -235 can provide 70 million watts of power through the nuclear motor for 1 day, but these power will be consumed in just 10 day, which is the most conservative estimate; If a quantum computer works for about 4 hours every day, its life span is only 2 years. If you work more than 6 hours, you may not even be able to stick to 1 year, which is also the most conservative estimate. Assuming that the quantum computer is 70 degrees Celsius per hour, the chassis reaches 200 degrees in 2 hours, and the radiator melts in 6 hours, this is still the most conservative estimate!
From this point of view, the quantum computer with high energy and short life is probably far from our life. The light of the future, let us wait and see! ~
Two: DNA biological computer
Scientists have found that deoxyribonucleic acid (DNA) has a feature that it can carry a large number of genetic materials possessed by various cells of organisms. Mathematicians, biologists, chemists and computer experts are all inspired by this, and are now cooperating to develop future DNA computers. The working principle of this DNA computer is based on instantaneous chemical reaction. Through the interaction with enzymes, the reaction process is coded by molecules, and a new DNA coding form is used to answer questions. 1995 reports for the first time that scientists have made a breakthrough in solving mathematical problems by "programming" DNA chains. Compared with ordinary computer, DNA computer has the advantage of small size, but it stores more information than any computer at present. The space it uses to store information is only trillions of times that of an ordinary computer. Its information can be stored in trillions of DNA strands. It only takes a few days for a DNA computer to complete any operation that all computers have done so far. Besides, it consumes only one billionth of the energy of an ordinary computer. The powerful function of DNA computer is that each chain itself is a microprocessor. Scientists can arrange 1000 billion chains in1000 grams of water, and each chain can be calculated independently. This means that DNA computers can "try" a large number of possible solutions at the same time. The electronic computer must calculate each solution from beginning to end until it tries the next one. Therefore, electronic computers and DNA computers are completely different. An electronic computer can complete many operations in one hour, but only one instruction operation can be completed at a time. It takes about one hour for a DNA computer to perform an operation, but it can calculate 65.438+0 billion instructions at a time. The function of the human brain is somewhere in between: about 10 trillion instructions are executed in one hour. DNA computer translates binary numbers into fragments of genetic code, and each fragment is a famous double helix chain. Scientists hope to decompose all possible patterns of DNA and put them into test tubes to make complementary digital chains, which will provide a basis for solving more complicated operations.
DNA ribozyme, a specific DNA structure, can be used to construct various DNA molecular logic gates, laying the foundation for the development of DNA computers.
DNA computing is a new research field developed by the combination of computer science and molecular biology.
According to the Chinese Academy of Sciences, Fan Chunhai, a researcher at the Shanghai Institute of Applied Physics of the Chinese Academy of Sciences, and Lin He, an academician of the Bio-X Center of Shanghai Jiaotong University, and Professor Zhizhou Zhang (now a professor at Tianjin University of Science and Technology) have successfully developed a new type of "DNA logic gate" by using DNA ribozyme through in-depth interdisciplinary cooperation, which laid the foundation for the development of DNA computers. Related research results have been published in the famous chemical magazine "German Applied Chemistry".
Because of the powerful parallel operation and super-high storage capacity of DNA molecules, DNA computing will probably solve some complex problems that are difficult to be completed by electronic computers, and it may also play an important role in the fields of drug delivery or gene analysis in vivo. Although DNA computing has great potential in the future, there are still many bottleneck technologies and basic problems to be solved, among which logic gates based on DNA molecules are the important foundation of DNA computing.
DNA ribozyme is a nucleic acid structure with specific enzyme activity screened by in vitro evolution. In this study, DNA ribozyme with DNA hydrolase activity was used. This hammerhead ribozyme can catalyze oxidation and cut the substrate DNA with the help of copper ions. Based on this DNA ribozyme structure, a DNA logic gate is developed by modular design. Input signals can be sensed by specific biomolecules to generate output signals, thus realizing logical judgments such as "yes" and "no", and can be combined into complex three-input logic gates "AND(A, NOT(B), NOT(C)". The combination of "not" and "and (a, not (b), not (c)" is a set of common operation symbols, so theoretically all operations of Turing machine can be realized by their combination.
The new feature of this logic gate system is to exclude the participation of RNA nucleoside in the previous design of DNA logic gates, and only use DNA molecules, thus avoiding the system instability caused by RNA nucleoside. Related research results have been published in the famous chemical magazine "German Applied Chemistry" (Angew) published in March. Chemistry. Edited by …, April 2006,1759.).
Three: Photonic Computers Existing computers use electrons to transmit and process information. Although the propagation speed of electric field in wires is faster than any means of transportation we have ever seen, from the point of view of developing high-speed computers, using electrons as the carrier of transmitting information can not meet the requirements of high speed, and the ability to improve the computing speed of computers is obviously limited. Photonic computers use photons instead of electrons to calculate, transmit and store data through optical fibers. In photonic computers, different wavelengths of light represent different data. By changing the states of electrons "0" and "1", it is far superior to the binary operation in electronic computers, and it can realize fast parallel processing of tasks with high complexity and large amount of calculation. Photonic computers will multiply the operation speed on the existing basis.
Disadvantages of electronic computers
[Edit this paragraph]
We know that any metal wire has resistance and capacitance. From the basic knowledge of electromagnetism, we also know that the combination of resistance and capacitance will produce "resistance" to electrons propagating in wires and reduce their propagation speed (the propagation speed is only about one thousandth of the propagation speed of light waves in vacuum), which is also called clock distortion. Because of these problems, the corresponding consequence is that electrons are "slow" to respond to rapid changes in the outside world. When the carrier frequency of information is high (that is, the signal changes rapidly), the electric signal transmitted on the wire can't keep up with the change of the information signal to be transmitted. As a result, the crosstalk performers read tongue twisters too fast, their tongues can't keep up, they mispronounce or go out of tune, the transmitted signals will be distorted and the computer operation will be wrong. Secondly, although the central processor of an electronic computer can process data quickly, the main memory can process a large amount of data. But all data signals must be transmitted through the bus. If the current density of the bus is too large, the electromagnetic interference will be great. Therefore, the electronic computer will also appear similar to the phenomenon that the speed of expressway intersection is limited because of its narrowness, and the speed of computer and operation is also limited. In addition, the integration density of integrated electronic devices used in computers is limited due to the interference of quantum effect. Theoretically, the highest integration density is 65.438+0 billion transistors per chip (the actual figure is much lower than this figure).
Advantages of Photonic Computer
[Edit this paragraph]
Instead of electrons, using photons as the carrier of information transmission may overcome the limitations mentioned above and make computers with better performance. Using photons as a carrier to transmit information has the following advantages:
1, photons have no charge, and there is no electromagnetic field interaction between them. In free space, several beams of light propagate in parallel, cross each other and do not interfere with each other. Thousands of beams of light can pass through an optical element at the same time without affecting each other. An optical system can provide 5 * 10 5-wire transmission information channels; Good quality lenses can provide information channels. If optical waveguides are used for transmission, optical waveguides can also cross each other, as long as their intersection angle is greater than about, there will be no obvious cross coupling. These properties are also called parallelism of optical signal transmission.
2. Photons have no static mass and can propagate in both vacuum and medium, and the propagation speed is much faster than that of electrons in wires (about 1000 times), that is to say, the information carried by photons is faster than that of electrons, and the interconnection between chips in a computer is not affected by electromagnetic interference, and the interconnection density can be very high. In free space interconnection, the number of wires per square millimeter area can reach 50,000, and if optical waveguides are used for interconnection, there can be 1 10,000. Therefore, using photons as the information processing carrier will create a computer with extremely high operation speed, which can theoretically reach 1000 billion times per second, and the information storage capacity can reach 10 18 bits. This kind of computer is called photonic computer.
3. Ultra-high speed operation. Photonic computer has strong parallel processing ability, so it has higher operation speed. The speed of electrons is 593km/s, while the speed of photons is 3× 10? 5km/s, for electronic computers, electrons are the carrier of information and can only be conducted through some insulated wires. Even in the best case, the speed of electrons in solids is far less than the speed of light. Although the operation speed of electronic computers is constantly improving, its capacity limit is still limited. In addition, with the continuous improvement of assembly density, the electromagnetic interaction between conductors will be enhanced, and the heat emitted will gradually increase, thus restricting the running speed of electronic computers; Photonic computers run much faster than electronic computers and have much lower requirements for environmental conditions than electronic computers.
4. Super-large-scale information storage capacity. Compared with electronic computers, photonic computers have huge information storage capacity. Photonic computer has an ideal light radiation source-laser. Photons can be transmitted without wires, and even if they intersect, there will be no interaction between them. In fact, the density of parallel channels for transmitting information without wires in photonic computers is infinite. The information transmission capacity of a nickel-sized mirror is many times that of the existing telephone cable channels all over the world.
5, low energy consumption, low fever, is an energy-saving product. The driving of photonic computers only needs a small part of the driving energy of electronic computers with the same specifications, which not only reduces the power consumption, but also greatly reduces the heat emitted by the machine, and also provides convenient conditions for the miniaturization and portable development of photonic computers. Scientists are trying to combine traditional electronic converters with photons to create a "hybrid" computer, which can not only process information faster, but also overcome the overheating problem in the operation of giant electronic computers.
Composition of Photonic Computer
[Edit this paragraph]
Photonic computer consists of optical components and optical mirrors, lenses, filters and other equipment. There are analog and digital photonic computers. The characteristic of analog photon computer is to directly use the two-dimensional properties of optical images, so the structure is relatively simple. This photonic computer has been used in satellite image processing and pattern recognition. The Star Wars plan proposed by the United States before is to develop this kind of computer for recognizing high-speed missile images. There are many structural schemes of digital photonic computers, two of which are considered to be of great development value. One is to adopt the mature structure in electronic computers, only replacing electronic logic elements with optical logic elements and replacing wired interconnection with photonic interconnection. The other is a new structure (optical neural network) based on parallel processing. In the 1980s, we made an optical information processor, and the digital optical processor was also successful. It consists of a laser, a lens and a prism. Although photonic computers have been successful, they can't catch up with electronic computers in function and operation speed. We mainly use electronic computers and will develop them in the future. However, from the perspective of development potential, it is obvious that photonic computers are much larger than electronic computers, especially in image processing, target recognition and artificial intelligence, and photonic computers will play a far greater role than electronic computers in the future.
Present situation of photonic computers
[Edit this paragraph]
Bell Laboratories announced the development of the world's first optical computer. It uses gallium arsenide optical switch, and the operation speed reaches 1 100 million times per second. Although this kind of optical computer is far from the theoretical one, it has shown great vitality. Humans have used optical cables to transmit data for more than 20 years, and the optical disc technology using optical signals to store information has also been widely used. However, in order to make a real photonic computer, it is necessary to develop a phototransistor that can control the change of another beam of light with one beam of light. Generally speaking, although scientists can realize such a device, the required conditions, such as temperature, are still harsh and it is difficult to enter the practical stage.
An optical technology company in Massachusetts, USA, Light Guide and Light Emitting Element System Company, is currently cooperating with NASA Marshall Space Center to develop an "optical" circuit board for manufacturing optical computers that control photon motion.
And it is expected to make a breakthrough this year. 1In May, 999, a research team of more than 20 people led by Singaporean scientist He, who works at Northwestern University in the United States, developed the world's smallest photon directional coupler by using nano-scale semiconductor lasers, which can decompose and control light in a semiconductor layer with a width of only 0.2 to 0.4 microns.