Основные закономерности эволюции компьютерных систем и сетей
Електронний архів E-archive DonNTU – (Electronic archive Donetsk National Technical University)
Переглянути архів Інформація| Поле | Співвідношення | |
| Title |
Основные закономерности эволюции компьютерных систем и сетей
Basic laws of computer systems and networks evolution |
|
| Contributor |
Аноприенко А.Я.
|
|
| Subject |
эволюция компьютерных систем
эволюция компьютерных сетей закономерности эволюции система закономерностей закон Мура эволюции сложности генома макрозакономерности мегазакономерности гиперзакономерности кодо-логическая эволюция |
|
| Description |
Basic laws of computer systems and networks evolution. Moore's law is the observation that the number of transistors on integrated circuits doubles approximately every two years. This empirical law was first formulated in 1965 as a law of doubling the number of active elements on the chip every year. In 1970, it got its name in honor of the author Gordon Moore – one of the founders of the company Intel. In 1975 this law was greatly improved: in the long term doubling the complexity of integrated circuits is possible only every 2 years – it's actually a modern form of the Moore's law. The analysis shows that the data accumulated to date, can significantly clarify and summarize these estimates. In particular, the list of the most productive Top500 supercomputers today is one of the most important sources for the analysis of patterns of growth performance of modern computer systems. Analysis of the Top500 list and other sources suggests that most indicators of development of modern computer systems increased by approximately one order every 4 years. These and other identified patterns allow us to construct a generalized model of the evolution of the performance of computer systems of different classes – a kind of periodic system of the basic laws of growth performance of different classes of computers. Four-year period, which is characterized by the growth performance for different classes of systems by approximately one order is the basis for the construction of the periodic system. By 2012, 10 such classes formed, but in fact it can be stated the emergence of a new class (which differs from the previous performance by an order) approximately every 4 years. At the same classes can be grouped in pairs in related groups, representing high-performance and relatively expensive (High-end), and mass (Low-end) kind of appropriate types of computers. Analysis of a variety of other data shows that in the development of information and computer technologies can be formulated as a whole system of hierarchically related laws, one of which is code-logical evolution. Рассматриваются основные закономерности развития компьютерных систем и сетей, которые более адекватно и всесторонне позволяют отслеживать и прогнозировать эволюцию средств и методов компьютинга. При этом, учитывая чрезвычайное многообразие различных данных и довольно частую их противоречивость, речь в первую очередь идет о разного рода моделях эволюции, отражающих основные качественные и количественные закономерности эволюции компьютерных систем и сетей с приемлемой и разумной степенью точности. |
|
| Date |
2014-08-02T08:07:47Z
2014-08-02T08:07:47Z 2013 |
|
| Type |
Article
|
|
| Identifier |
Аноприенко А.Я. Основные закономерности эволюции компьютерных систем и сетей // Научные труды Донецкого национального технического университета. Серия «Проблемы моделирования и автоматизации проектирования» (МАП- 2013). Выпуск № 1 (12) – 2 (13): Донецк: ДонНТУ, — 2013. С. 10–32.
http://ea.donntu.edu.ua/handle/123456789/27275 |
|
| Language |
other
|
|
| Publisher |
Донецкий национальный технический университет (ДонНТУ)
|
|