Silicon’s limitations set against Moore's law
Integrated circuits or chips are the heart of all computers. These chips are largely made from silicon, the second most abundant element found in the earth's crust. Silicon can be found every where on this earth, such as in sands or in the glasses. Perhaps the most revolutionary use of silicon is in computer chips that changed the method of computing and the way we live.
Silicon enabled the engineers to integrate more components onto the same size chip, almost doubling the number of components on a given piece of silicon roughly every two years. But there is a limit on the stretchibility of the silicon chips. As the individual components on a chip get smaller, the physical possibility of doing so is reaching to an end.
However the chip manufacturers have devised new ways to cram more number of smaller transistors onto a single chip. By setting the transistors in different arrangements, engineers have been able to create a circuit that can store a value or perform a calculations.
Constraints at atomic level
The number of transistors on chips keep increasing at a double rate in every two years. This trend was predicted by Intel founder Gordon Moore in 1965 and henced named Moore's Law. To integrate more number of transistors on a single chip, the size of transistors has to keep reducing. Lately the smallest transistor that could be placed on a chip was 65 nanometers across. But to cope with Moore’s Law humming, even 65 nm isn’t enough. Some of the companies are now producing chips based on 45 nm devices.
Atoms and molecules are the building blocks of the transistors and their dimensions just cannot be reduced. In case the transistors or their components continue to get smaller, a point will reach where the placement of individual atoms will affect their behavior.
In order to work properly, the thickness of the silicon layers in chips needs to shrink proportionally to the length and width. At the 90 and 65 nm horizontal sizes, the thickness of “gate oxide” layer, which acts as an electrical insulator between conductive layers, remains only 1.2 nm. It sums up to the thickness of five individual atoms. At this level it is ok. But the problem is that at 45 nm, the gate oxide would have to be even thinner. In such a case electrons would start tunneling through it, ruining its properties as an insulator. Intel worked around this problem by using a new layer based on the element hafnium.
What after silicon?
Chipmakers and scientists have been discussing about the alternative forms of computing technologies to keep Moore’s Law alive. The idea which comes foremost is Optical computing, that would use photons instead of electrons. But the optical technology works best to connect processors together over a distance, rather than inside the chips themselves.
Quantum computing is another alternative that uses the attributes of elementary particles such as electrons as the basis for calculation. Unlike traditional digital computing, in quantum computing a bit of data can be stored as 1 or a 0 both at once. However the good news for Moore’s Law is that it seems healthy for at least another 10 years or so.
No comments yet.