HISTORY OF TECHNOLOGY, TECHNOLOGY, OBJECTS AROUND US
Neurocomputer. History of invention and production Directory / The history of technology, technology, objects around us A neurocomputer is an information processing device based on the principles of operation of natural neural systems. These principles were formalized, which made it possible to talk about the theory of artificial neural networks. The problem of neurocomputers is to build real physical devices, which will allow not only to simulate artificial neural networks on a conventional computer, but to change the principles of computer operation in such a way that it will be possible to say that they work in accordance with the theory of artificial neural networks. The terms neurocybernetics, neuroinformatics, neurocomputers came into scientific use recently - in the mid-80s of the XX century. However, electronic and biological brains have been constantly compared throughout the history of computing. N. Wiener's famous book "Cybernetics" (1948) has the subtitle "Control and Communication in Animals and Machines". The first neurocomputers were Rosenblatt's perceptrons: Mark-1 (1958) and Tobermory (1961-1967), as well as Adalin, developed by Widrow and Hoff (1960) based on the delta rule (Widrow's formula). Adaline (Widrow learning adaptive adder) is now a standard feature in many signal processing and communication systems. In the same series of the first neurocomputers is the program "Kora", developed in 1961 under the leadership of M. M. Bongard.
Numerous elements (devices) of a computer located in its system unit can be divided into only five main groups. These are the central processing unit, memory, bus, power supply and numerous analog-to-digital and digital-to-analog converters (ADC and DAC). The processor is directly connected to the elements of fast (random) memory. It is also called random access memory (RAM) or random access memory. When you turn off the computer's power, it is cleared and all data in it is lost. Data is retained in non-volatile memory even after the computer is turned off. Most often, it is larger in volume than RAM, although not as fast. These are hard, floppy and optical disks, magnetic tape, etc. Data is transferred via the bus between the devices of the system unit. ADCs and DACs convert information from analog to digital form: into sets of numbers, usually binary, and vice versa. ADC and DAC are called controllers. Any controller contains a microprocessor, which means that it is a computer, but not a universal one, in which it is installed, but a specialized one. Programs are “soldered” in the microcircuits that are executed when the computer is turned on and, as it were, enliven it, turning many parts connected by wires into a single whole - into a universal information converter ready for use. Microprocessor technology is already approaching fundamental limits. Gordon Moore's predictive law states that the density of transistors in a microcircuit doubles every year and a half. Surprisingly, for the past twenty years it has been carried out. However, following this law, by 2010-2020 the size of the transistor should be reduced to four or five atoms. Many alternatives are being considered. Technologies capable of exponentially increasing the processing power of computers include molecular or atomic technologies; DNA and other biological materials; three-dimensional technologies; technologies based on photons instead of electrons, and, finally, quantum technologies that use elementary particles. In the XNUMXst century, computing technology will merge not only with communications and mechanical engineering, but also with biological processes, which will open up opportunities such as the creation of artificial implants, intelligent tissues, intelligent machines, "live" computers and human-machine hybrids. Today, one of the most promising areas in microelectronics is neurocomputers. Their device, or architecture, is different from that of conventional computers. Microcircuits are similar in structure to the neural networks of the human brain. This is where the name came from. Hence the features of the neurocomputer. He is capable of learning, which means that he is able to cope with tasks that a conventional computer cannot do. His main trump card is solving problems without a clear algorithm or with huge flows of information. Therefore, neurocomputers are already used today in financial exchanges, where they help predict fluctuations in the exchange rate and stocks. It is clear that the military did not stand aside. Neurocomputers, recognizing images, correct the flight of missiles along a given route. The neurocomputer has become, in essence, the flag of a new wave of research and development in the field of neural network methods of information processing, almost completely replacing the term "neurocybernetics". The hopes associated with early work on the creation of artificial intelligence (AI) systems were naturally transferred to neurocomputers, which were understood in a broad sense as prototypes of an "artificial brain" - an intelligent system that should be built and function similarly to the human brain. The prefix "neuro" emphasized the difference between such a system and a traditional computer and functional proximity to the brain.
The real state of affairs rather quickly forced us to narrow the understanding of the term "neurocomputer" to identification with artificial neural networks. In most modern works, this term (or the term "neurocomputing") is used to refer to the entire spectrum of work within the framework of an approach to building AI systems based on modeling the elements, structures, interactions, and functions of various levels of the nervous system. In the modern sense, a neurocomputer is a specialized software or hardware-implemented computing device that simulates the operation of a neural network. The first hardware neurocomputer in the USSR was developed in 1988-1989. based on the ideology of ensemble stochastic neural networks. The work was carried out under the guidance of Doctor of Technical Sciences. E.M.Kussulya, to whom by that time Nikolai Mikhailovich had already transferred the Department. The first model of a neurocomputer (1989) was created on the domestic element base and was a prefix to a personal computer. In subsequent layouts, a more advanced element base was used. In 1992, together with the Japanese company WACOM, the latest version of the neurocomputer was developed and experimentally tested on pattern recognition problems.
The subsequent work of the Department was related to the development of neural network information technologies. Efficient neural network classifiers were created that were used in the tasks of texture recognition, identification of a person by voice, recognition of handwritten characters, continuous written words, etc. Despite the applied nature of these works, the Department retained the global approach to the problems of artificial intelligence instilled by N.M. Amosov, the ability to see the problem as a whole and accumulate experience for the next "breakthroughs". Author: Musskiy S.A. We recommend interesting articles Section The history of technology, technology, objects around us: ▪ Compass ▪ Elevator See other articles Section The history of technology, technology, objects around us. Read and write useful comments on this article. Latest news of science and technology, new electronics: Artificial leather for touch emulation
15.04.2024 Petgugu Global cat litter
15.04.2024 The attractiveness of caring men
14.04.2024
Other interesting news: ▪ New OLED displays can be folded over 100000 times ▪ SPARC T5 servers powered by the world's fastest microprocessors ▪ Spiders, feeding on graphene, weave the strongest web News feed of science and technology, new electronics
Interesting materials of the Free Technical Library: ▪ section of the site Lecture notes, cheat sheets. Selection of articles ▪ article From the ship to the ball. Popular expression ▪ article Economics and Finance. Big encyclopedia for children and adults ▪ article Electric knife. home workshop ▪ article Flashing red light. Encyclopedia of radio electronics and electrical engineering
Leave your comment on this article: All languages of this page Home page | Library | Articles | Website map | Site Reviews www.diagram.com.ua |