Thursday, June 10, 2010

How Fast Is 4 Hp Motor

Silicon dead, long live the neochips! Nonsense

Reading
few weeks ago the English publication of "Scientific American "," Scientific American ", an article caught my interest. It was called" Microchips in the immediate future "and was the same on developments and research that are developing throughout the world to create new microprocessors, I quote:

"Innovations in design and manufacture of semiconductors have enabled it to meet the famous prediction by Gordon Moore in 1975, doubling every two years the number of transistors integrated on a silicon microchip. As miniaturization progresses, the electrical signals have to travel shorter distances and operation is accelerated, which significantly reduces industrial costs and prices to the customer. For 35 years has remained fairly accurately that path.

knew, however, that at some point reach the limit. When the thickness of the transistors is reduced to a few tens of atoms, as encountered with the basic laws of physics. Before even going to practical problems arise. Closely transistors so tiny crowd while maintaining a high production yield - may útiles/defectuosos-- chips would cost too much. In addition, heat generated by the operation of the tangle of transistors could start Cook the elements themselves.

Of course, the barriers were raised several years ago. The main reason that the current personal computers today use dual-core tablet with two processors instead of one (chips "dual core") are the problems created by the cooling of a large number of transistors encapsulated in a single chip. Engineers prefer to juxtapose two or more chips and schedule them to process information in parallel.

seems, then, that Moore's Law is becoming outdated. How can then build more powerful chips? Could be chosen or refine architectures other nanomaterials likely to be assembled atom by atom. Another way would be to refine new methods for processing information, such as quantum and biological computing. Below are a range of innovations, many of them in the experimental stage, that in twenty years could maintain software products in the path of the "smaller, faster, cheaper", which has served us well.
Projects are more likely to come out winners in this race for the new processor architecture are, according to the magazine:


  • The Memristor of Hewlett-Packard : Today transistors are smaller than 32 nanometers wide, smaller sizes get through the lithographic process is very difficult. Apparently, one of HP's solutions to this problem is to design the transistor with cross bars, where smaller sizes are not obtained but if more computing power. So "instead of making all the transistors on the same plane (as cars crammed into the lanes of a highway congested silicon), two separate sets of parallel nanowires will be crossed at right angles on different planes, including a separation layer is interposed Gros of a molecule. "
  • Heating: Refrigerated or ventilation The passage of electric charges through the circuit has the disadvantage of being a source of heat which can end up burning their own circuits. Is what is known as micro-electronics the problem of heat dissipation of the circuits. In computers the problem is solved by placing a fan in some new tools have been used for polished aluminum cases that help dissipate the heat to circulate liquids through a micro channel. Intel is testing a solution in half the chips, "a lattice in the form of thin film of bismuth telluride in the package that contains the chip. That makes thermoelectric material gradients temperature in power and cooling the chip. "Another way is that developed by Ventiv , which attempts to place a tiny fan in solid form, which needs no blades, and will create a breeze through a ventilation effect crown.
  • multiple cores: Intel is developing processors like the Intel i7 , a chip has four cores working in parallel, allowing them to "process more data at a given time, less power consumption and less heat generation. "The challenge however for this type of solution that is in parallel programming languages are not very advanced in regard to consumer applications.
  • and self-assembling nanotubes: Based on developments in nanotechnology, companies like IBM are investigating the use of carbon nanotubes , as is the case of an old semiconductor (CMOS complementary metal-oxide) which is conductive substrate just a carbon nanotube instead of the classical silicon. But how to make molecules or particles are placed where they (the assembly), since a path can come from what is known as autoemsablaje, where they themselves are sorted according to a schedule set. "IBM has demonstrated how to make memory circuits using polymers linked by chemical bonds. By revolver on the surface of a silicon wafer and then heated, the molecules are stretched and form a honeycomb structure whose pores measure only 20 nanometers wide. Then these settings can be recorded in the silicon, which results in a memory chip that size. "
  • ultrathin graphene: A way to reduce heat transmission speed is to shorten the path that electrical impulses must go. Graphene, like a web of metal but the thickness of an atom, electrons can be transmitted faster than any other material.
  • Computer Optical Based on the use of laser light in the task of transmitting electricity. Intel is trying to develop an optical chip "fitted with a controllable light source inside [that] can compute quickly. Electrons and holes in the layers of indium phosphide in the center recombine to generate light, which spreads down waveguide and silicon through a layer of glass. " The researchers were able to combine the properties of light-emitting indium phosphide with the routing capabilities of silicon light within a single hybrid chip. When voltage is applied, light generated in the Indium Phosphide enters the waveguide of silicon, creating a continuous laser beam que puede ser utilizado para conducir otros dispositivos fotónicos de silicio. El láser basado en silicio puede ofrecer un uso más amplio de la fotónica en ordenadores porque el coste puede ser reducido de forma muy importante al usar técnicas de fabricación de silicio a gran escala.
  • Lógica orgánica: Un transistor de los de toda la vida tiene una forma lógica en Y, donde el tronco de la Y es la puerta de entrada, en la cual al aplicar un voltaje, provoca la circulación de electrones entre la fuente y el drenador (las ramificaciones de la Y). En teoría, lo que se opina es que las moléculas que tengan la misma arquitectura o estructura, pueden servir de transistores, y teniendo en cuenta el tamaño some of them tiny, it would reduce significantly the size of the chips.
  • Quantum computing: The bit is dead, long live the qubit. In its day, I wrote an entry on the subject.
  • biological Computer: Based on replacing "transistors by structures normally found in living organisms. [...] A biological chip also enclose a number of elements by several orders of magnitude larger than the regular chip, parallel processing could provide a large scale. " Is currently studying at the Weizmann Institute of Science the use of a processor through DNA.
The future of new technologies is proceeding at full speed and find the company that it will cause a quantum leap as assumed at the time the silicon chip. Despite how excited I am, I can not help feeling sad about it, seeing that none of the ideas proposed in this paper we have developed in our country. This is the end of the day, to be part of what will be the spearhead of the new computer and everything to cause or be members of the caboose. And unfortunately, our politicians seem more concerned and occupied with other things to devote more time to provide ourselves with competitive advantages in lands that lie ahead.

0 comments:

Post a Comment