|
|
|
|
|
Transistor Minimization Hits Dead EndOld Invention, New Problem, No Solution |
By Gabriel Gache, Science News Editor
17th of December 2007, 09:02 GMT
Adjust text size:  
|
| |
Co-founder of the Intel Corporation, Gordon E. Moore, predicted in 1965 in a paper that the number of transistors that can be inexpensively placed on an integrated circuit would grow exponentially and duble every two years. Since then this prediction has become known as Moore's Law and described relatively accurately the advance of technology in the 21st century.
However, 60 years after the invention of the first semiconductor transistor and about half a century since the creation of the first integrated silicon chip, Moore's Law seems to get to the end of its prediction streak. The computing power of a silicon microprocessor is directly related to the number of transistors packed inside it. But the miniaturization of the transistor cannot go forever and will reach a point in the future where it will not be economically viable to place more transistors on a silicon chip to create even more powerful computers.
The problem is generated by physical limitations, as manufacturers squeezing more transistors on a silicon chip of the same size would not be able to produce processors that will provide a dramatic gain in performance at low production costs. In order to avoid such a catastrophy microprocessor manufacturers are investing even more money into developing technologies, in the hope that they will be able to produce a device that would replace the semiconductor transistor and start a second digital revolution.
Alternative technologies already using quantum spin manipulation and optical chips are in the developing stages. The first semiconductor transistor, the field effect transistor, was described by a German physician in the middle of the 1930s, for which he has obtained a patent but he never got the chance to build it. Twelve years later, in late 1947, the first bipolar transistor was successfully tested at the Bell Laboratories. Transistors are electronic components, made of two p-n semiconductor junctions, formed by three separate alternating layers of semiconductor, which are used to switch electric signals between two terminals by controlling the current through a third one.
Due to this particular configuration, transistors can only process binary signals, which greatly reduces the computing power. Alternative devices that could process multiple signals have been invented but proved to be too unstable in providing accurate signals to be used in computer processors.
As they where invented, transistors quickly became the ideal replacer for the bulky vacuum tubes that where previously used for amplifying and switching electrical currents. Recognizing their potential for developing incredibly small electronic circuits, Texas Instruments Inc. representative Jack Kilby and Intel Co-founder Rober Noyce, simultaneously invented in the late 1950s the first integrated circuit for which they received the Nobel Prize, and opened the way to the creation of the first microprocessor, currently used in most of the electronics today.
Nevertheless, though it was first a component that triggered the digital age, the transistor seems to block future technological advance. The biggest problem presented by minimization is the fact that, as components become thinner, more heat escapes from the chip, making it extremelly inefficient. Furthermore, components packed closer together have a better chance of leaking electric currents between them.
The solution would be to replace the silicon dioxide insulator in the field effect transistors with a material having improved insulating properties that would increase the transistor's performance.
|
|
| Rating: |
|
Good (3.4/5) |
5 vote(s) so far |
|
|
|
|
User opinions: |
 No user comments yet.  Be the first to express your opinion using the form below! |
|
|
