3D Transistors


  3D Transistors[1].doc (Size: 230 KB / Downloads: 3)

Introduction

As the technology grows, in the field of processing and transistors, the size of the devices, hence transistors need to decrease to meet the requirement of the technology. Devices that everyone uses daily such as mobile phones and laptops are moving in a direction to smaller size and more efficiency. If we take laptop as an example, one of the most important requirements is the battery lasts longer. Another aspect is the small size of laptops to be convenient to take them and use them as a mobile device. Smaller laptops will require smaller parts to put in them.
As the size of transistor gets down to 45nm, as it was discussed in previous sections, leakage current becomes a big issue. Because of the small size of the drain-source channel is so small, it is hard to control the amount of off-current flowing through the channel.
In2002 Intel introduced a new generation of transistor claiming less power consumption. As it was discussed in the previous section, strained silicon and High-K material can help us to overcome this problem for now, but as the technology is growing, the new generation will become handy. Called “3D transistors” use more than one gate to control the current flowing from drain to source in the “on-state” of the transistor.


Leakage current and power consumption


When the transistor is at off-state there must be no current flowing down the channel. As it was stated before when transistors shrink, which basically means smaller size of channel gate, leakage current becomes an issue. The resistance that gate creates when there is no voltage applied, prevents the current from flowing, making the transistor off. When the channel is too short the resistance loses its effect and there will be current from drain to source.


3D transistors in processors

As for computers, although the solutions that have been explained and discussed in “Current technology” section of this report may help us to move to the new generation of fast processing, but new technologies will be required as the technology grows and faster processor will be needed for different applications.
As it was explained in previous sections, most parts of processor are made of transistors. One of the factors of fast processors is faster transistor.
As the transistor gets smaller the length of the drain-source channel decreases, so the time taken to current travels through the channel will be less, therefore the transistor will switches states faster[5], hence the processor will be faster.
As the size of the transistor gets smaller, the length of the channel decreases. Although this causes some problems, which the new design deal with them, but one advantage of short channel length is that the transistor can be switched on and off faster.
Another advantage of three dimensional transistors over single gates is the fact that because of the 3D design and using the 3rd dimension, the new design will use less space on the chip. This means more transistors can be placed on one chip and hence Moore’s law will still be around for another generation.



More designs
The term “Tri-gate” is used by Intel for three-dimensional transistors. There are other different designs. Intel tri-gate transistor, as the name suggests, deploys three gates instead of one.
Other designs use the same concept as that has been explained. The difference between them is the performance and how much each design can help to improve today’s technology and overcome problems that planar transistors are facing.
AMD with corporation of University of California have developed a transistor with two gates. FinFET also uses ultra thin silicon to reduce the amount of leakage current in the off-state of transistor. This improvement enables to fabricate transistor on smaller chips with better computational performance. [9]