The fifth generation of computers 1982 - our days

The fifth generation of computers is a government program in Japan for the development of computing technology and artificial intelligence . If we talk about previous generations, the first is lamp computers, the second is transistor, the third is integrated circuits, the fourth is microprocessors. But the fifth generation is not related to this gradation. Like the previous generation. The fifth generation of computers is the name of the “action plan” for the development of the IT industry. And despite the fact that the fifth generation is based on microprocessors as the fourth one, i.e. they have a common element base. Namely, by this criterion, computers are divided into generations. Nevertheless, today's computers belong to the fifth generation.

Japan began its large-scale program in the early 80s. Their goal is not to change the element base of computers. And to change and improve technical approaches, programming methods and develop a scientific direction in the field of artificial intelligence. At the beginning of its project, Japan has invested half a billion US dollars. At that time, it was not as technically developed as the USA, Europe. The Ministry of International Trade and Industry of Japan has set a clear goal - to get into the lead. It was at that time that the term "fifth generation of computers" was born. Computers of the fifth generation must achieve superconductivity and a huge number of processors on one substrate must be integrated into them.

The Japanese Information Development and Processing Center has set goals. The main one is the development of technologies for the logical processing of knowledge, one of the leading directions of artificial intelligence. Creating workstations with high performance and distributed functions. Creation of fifth-generation supercomputers for scientific computing that will operate with huge databases and knowledge bases.

Even then, there was a need for computers with processors operating in parallel and extracting data from structurally constructed databases using the interface of high-level logical languages. The performance of such machines should reach one billion logical conclusions per second. It is necessary to create such a computing environment, which inside will create its own virtual environment, depending on the task.

One of the ways to improve the performance of the fifth-generation computer is to implement software solutions at the hardware level. Scientific advances in artificial intelligence. It is necessary to translate into a practical base. This is a computer typing dictation with speech recognition. And a software translator from language to language. Programmatically determine the meaning of the text for a decision on which rubric should be placed in it. Super computers should solve the problem of mass use.

Japan planned to complete this project in 10 years. And by the beginning of the 90s to reach a new level of technical development. At that time, Japan firmly conquered the consumer electronics and automotive industry, which worried the United States very much. In response, the Americans began to develop their own parallel computing programs. The largest projects involved the American Corporation for Microelectronics and Computer Technology (MCC). Europe is confident in the future of parallel computing. Begins plans in this industry British company Alvey.

In the Soviet Union, they attempted to keep up with their Western colleagues. There was a desire to create a fifth-generation prototype computer. For the future multiprocessor computer, they invented the bright name "MARS". But even then the gap with the Japanese, in the field of microelectronics, was 10-15 years. The whole project was based on old engineering solutions. And obsolete programming languages ​​like Modula-2. It was possible to create a multiprocessor computer "MARS". This was his only difference from the rest of the computer. And this machine did not meet the definition: “fifth generation computer”.

However, the implementation of the project “computer of the fifth generation” turned out to be more complicated than it was originally intended and not feasible in ten years. The functional language of the Prolog was chosen as the basic language for the fifth generation computer. But he did not support parallel computing. His work in a multiprocessor environment was not effective. Despite all the attempts to modernize it. It was decided to create new types of programming languages. This task was very difficult. Software development corporations have proposed new languages. But each of them had significant shortcomings. That did not allow the full use of parallel computing.

There were also hardware difficulties to create a fifth-generation computer. Technical development quickly overcame the difficulties that were considered not feasible before the start of the project. Parallel operation of several processors did not give the high performance, which was originally expected. Developed in the laboratory machine. Quickly obsolete. There were commercial computers that were already superior in speed. The project called the “fifth generation computer” was not successful. Because The development of information technology has taken a different path.

A graphical user interface has appeared. Which was not originally provided for in computers of the fifth generation. The Internet has appeared, which has changed the understanding of the structure of storage and processing of information. Search engines have evolved that used new data processing methods.

At the beginning of the project, the "fifth generation of computers" planned a complete transition to logical programming languages. Type Prolog. But they did not justify themselves. The main idea was to make the system learn. The system reached its peak, and then fell. And all the training needed to be repeated.

Plans to replace software with their hardware counterparts, was unsuccessful. Engineers had such a vision in the previous generation of computers. But today. The situation has changed radically. The process of development of information technology has gone the opposite way. The hardware was simplified. Providing versatility. And all tasks were shifted to the software.

The ideas of Japanese scientists were not successful. Initially, the development vector of computers was not correctly chosen. The hardware development perspective has been underestimated. Opportunities in the development of artificial intelligence were overrated. This area was more difficult than expected. Many theoretical developments in this field have not found their practical application. Artificial intelligence has not gone beyond academic tasks. Many methods have remained fun scientists.