The computer equipment future

Appearance of computing circuits was a natural stage of development of a civilisation at its industrial stage. It is linked by that a science, engineering and the industry, and also defensive tasks demanded fast processing of considerable streams of the information. At the initial stage the problem sharpness was removed by the information centres created on the basis of difficult, powerful (on those times) and expensive computer complexes serviced by the big staff of specially trained experts.

However the present break in data processing has occurred after appearance hardly more two tens years ago the personal computers created on the basis of rather cheap components. It has allowed to create a considerable quantity of computing circuits that has provided transition from the centralised way of information processing to the decentralised. Thanks to appearance of tens millions computers there was a reality their wide implantation practically in all spheres of human activity. As the result has substantially varied work of hundreds millions people, but it is much more important - public priorities of development have varied. The information became the object of work both extremely valuable and important product. In the developed countries to 90 % of riches it is linked to its processing, accumulation and sale. Efficiency of this activity are substantially defined by functionality of computer resources which in turn depend on a level of development of high technologies. All it also became priority directions of development of a modern society.

So, the mankind has entered an epoch of an informational civilisation from it yet not learnt laws of development, and also lockups, breaks and dangers.

It is necessary to mark, that the mankind always tried to glance in the future to accelerate arrival of the light and rich future, and also to avoid possible errors and annoying lockups. For coordination of the efforts of a policy, scientific, engineers and technicians try to co-ordinate the creative activity. Numerous forecasts and development estimations are this noble purpose. The contribution to this business is brought also by regular sessions IDF (IDF — Intel Developer Forum), devoted to high technologies in the field of computers and electronics. Substances of this forum can form a good basis for short-term and long-term forecasts of development of the computer industries influencing, by the way, directly on a level of development of a society and a life of its members.

At meetings IDF led in the different countries of the world within all year, the advanced processing techniques and products of electronic branches of the industry defining possibilities of a modern computer technology are discussed. Besides, at these meetings by the largest experts perspective directions are discussed, and also possibilities of existing and future hi-tech processing techniques are parsed.

And though already many it is accepted to decrypt the first character in IDF as Industry and to consider the Forum as action of all computer industry, it is impossible to forget, that Intel corporation is the in the lead corporation substantially defining development of all branch.

Key to the described future, according to Patrick Gelsingera, the vice-president and the Main Technical Director of corporation of Intel (Patrick Gelsinger, Vice President, Chief Technology Officer, Intel Corporation), is Moore's well-known law. Many leading computer experts adhere to the same opinion also.

More three tens years ago Gordon Moore (Gordon Moore, Chairman Emeritus of Intel Corporation) on the basis of the analysis of development of monolithic microcircuits has predicted regular doubling of number of transistors in difficult electronic arrangements. Originally doubling was predicted every year, and the subsequent some years this estimation wonderfully corresponded to real complicating of integral circuitry. However further this prediction has been modified: doubling of number of transistors in processors occurs each 1,5-2 years. The given estimation saves the urgency already almost four ten years. It is named now Moore's as law (Moore's Law).

Gordon Moore, the author of the law of Moore (Moore's Law)

Moore's law became a rule for experts and the businessmen, allowing to predict potential possibilities future, yet not created products.

At the analysis of possibilities of current technologies there is a question on periods of validity of the law of Moore. How long computer chips will evolve on an exponential low? Whether there is a limit to ambitions of the implementators increasing quantity of transistors in chips at constant reduction of the sizes of units, making their basis?

Really, in 80th years experts in semiconducting units dreamt of processing technique 1 microns, and argued on periods of reaching and possibility of overcoming of this boundary. Today speech already goes about barrier overcomings to 0,01 microns (10 nanometers). When as a result of reduction of the sizes of units there will occur permanently removed limit? And when, at last, Moore's law will cease to operate?

The head of corporation of Intel Krejg Barret considers, that development of solid-state technologies will allow to realise the processors working on clock rate of 30 GHz

Nevertheless, laying aside frequency parametres of computer accessories, it is necessary to recognise, that limits of reduction of the sizes of semiconducting units, certainly, exist. It follows from nature laws in which basis the digital structure of the world environing us lays, including an atomic structure of semiconductors. However to limits, comparable atoms with the sizes still rather far. For this reason prematurely to estimate the future architecture of the computer components which basis according to the forecast will be made by billions semiconducting units.

Patrick Gelsinger reflects on the future of processing techniques

As to perspectives and periods of validity of the law of Moore their Patrick Gelsinger in one of the reports on IDF here is how has estimated: “Someone speaks 10 years, 15 years. I promise to you, that while I will not retire, at least, 2,5 decades, Moore's law will exist. The future never was for us such clear as it is now. We will move forwards together with Moore's linked to development of business and processing technique the law. And the end while out of sight”.

In process of development of electronic and computer processing techniques complexity and productivity of processors permanently raises. On this path there was a set of objective obstacles, but they were step-by-step overcome by scientists and engineers at the expense of creation of new architectures and perfection of the newest processing techniques. And at Forums the copies of processors working on frequencies, exceeding 5 GHz were already shown.

Here it is necessary to remind, that at the heart of overwhelming quantity of chips released now usage of traditional field MOS devices lays (MOS-processing technique field-effect transisors — Metal-oxide-semiconductor contain on three contact electrodes-areas — the Source, the Drain, the Lock; the current a transistor Source-drain is controlled by the power submitted on a lock; the current through the Lock is practically equal to null). CMOS transistors — complementary MOS devices became development of these units . Perfection of solid-state technologies allows to reduce their sizes.

The architecture traditional and teragertsevogo the transistor

However reduction of the sizes of semiconducting units and high operating frequencies generate variety of problems. For example, experts of Intel have already shown the transistors containing units which thickness made only three atoms. At so mesh sizes it is practically impossible to eliminate uncontrollable currents of leakage flow. Moreover, quantum effects, for example, the Esaki effect providing penetration elektronov through potential barrier of insulating layers already strongly affect. The uncontrollable currents of leakage flow multiplied on number of used transistors, can reach considerable values. Besides, on the high frequencies reaching sometimes of tens gigacycle, even micron explorers can bring undesirable delays of signals in sort of presence of own inductivity. Allocated in immediate proximity from explorers and units, they as, however, all components of an inner pattern, create the mutual spurious capacitances influencing passing of signals, aspiring to break stability of operation. Besides, these of inductivity and size raise energopotreblenie and teploobrazovanie.

Article continuation>>

To pass to razrelu Processors