As some say, the greatest discovery to mankind has been fire. Well, you may argue it might be penicillin or fungus in general. However, in the current age of technology, something that is precious to people more than anything is semiconductors, more precisely, transistors. Whenever you hear the word ‘transistor’, you may picture an IC or an electrical component with three pins hanging out of it. You are not wrong. Transistors are the heart of today’s technological advancement. Since its inception in 1947, its development has significantly shaped how we see and perceive the world today. Numerous medical advancements, and military developments, from your light bulbs to the nukes, anything electronic you associate today with, may have a transistor. Talk about the numbers? Well, you might just have a few billion on the top of your palms and even more on top of your lap.
Since the development of the first microprocessor in 1968, the growth of the chip industry has been exponential. By the end of the year 2021, the global market size of the semiconductor industry was around USD 527.88 Bn which was expected to touch the figures of $ 575 Bn. As much as the market size is growing, the chip itself is getting smaller and smaller. It has reached such a minuscule size that the labs where these are designed and manufactured are more secure and meticulously configured than a path lab.
Looking back at the development of transistors, the first transistors to be made in the late ’40s were typically a centimeter in length. Soon they were in the millimeter scale. Gordon Moore, the founder of the famous Intel Company, saw and put out a statement that ‘the number of transistors in an Integrated Circuit (IC) doubles every two years.’ If we were to put that to simple terms and access theoretically, the number of transistors in IC today should be around 248. However, the current A16 Bionic processor of an iPhone contains over 16 billion transistors, that are manufactured in a “4nm” process at Taiwan Semiconductor Manufacturing Company Limited (TSMC). This is mainly because of the enhancements made in material quality and resulting increased efficiency. To put it into perspective, a normal silicon atom has a diameter of 0.21 nm. This means we are dealing with a transistor that is roughly the size of 20 silicon atoms. This level of design confronts intricacies of a scale never encountered before. And talking about intricacies, something very amusing awaits in that nanoscopic world.
Ever heard of the word ‘quantum’? Definitely. Let’s be a little more familiar. Schrödinger’s cat? For sure. Well, then why not take a little dive into it?
You see, the physics we learn today has been through a lot of evolution. From thinking that the sun revolves around the earth to discovering traces of methane and water in celestial in different galaxies, science has come a long way. Lot of things that we learn today is based on Newtonian Physics, or some say classical physics. Even light was considered as traveling in a medium called ‘ether’ until in the late 19th century, Faraday designed the cathode ray tube. Soon, scientists started to notice phenomena that simply couldn’t be described with the help of Newtonian Physics. The denial of such a series of laboratory experiments to be explained by the prevalent classical mechanics, electromagnetism, and thermodynamics led to multiple questions. The answer to that would come up years later when Max Planck proposed a statement, a rather radical assumption. After seeing such peculiar behavior from the experiment on black body radiation, he concluded that
the atoms do not give radiation away continuously, rather in a discrete manner in multiples of fundamental amounts.
The famous equation ‘e = hf’ is the equation that defines those ‘fundamental amounts’, where e is energy, f is the frequency of radiation and h is the Planck’s constant. These fundamental amounts are now what we call “quantum” or simply, “quanta”.
Whenever we discuss a macroscopic physical phenomenon, we investigate properties such as pressure, atmosphere, temperatures and many more. However, as we go down to the nanoscale, things like electrons and their moments are studied meticulously. A single electron has a lot to offer when we want to understand the working of a nano transistor. Now coming to the point, we all have heard of the famous Schrödinger’s cat. It is said to be both alive and dead at the same time, but how is it even possible? Well, it’s just an analogy to make you understand the principle of quantum superposition. When we are talking about an electron revolving around a nucleus in an atom, we don’t know its exact location until we know its exact location. You see what I did, right? In other words, the electron is in the state of superposition. It has no exact location around the atom. It’s omnipresent at a certain energy level. Let’s call that energy level a potential barrier. You could know that as an orbital. The position of that electron is given in a probabilistic figure by the wave function. Wave function provides measurements on particles and their positions which can’t be determined as their real-world equivalent. These are discrete measurements that are the function of space and of their time. So whenever we make an observation or take a measurement, the wave function collapses into a single state. That single state is the determinant of that instance in time and space. A nanosecond later, the measurement could have produced a different collapse of the wave function and a very different observation would have been made.
Now we reach our problem. When we arrive at a scale where transistors are so close to the size of an atom. We are pushing the nanometer barrier, where the next-gen transistors have a gate length of a nanometer. At this level, we observe a very interesting phenomenon. As we understand, the electrons exist in a probabilistic manner around the atom and the working principle of a transistor is the movement of an electron from its gate. Now imagine, what if the electrons start moving without even you, the computer, or let’s say the device providing any electrical charge to it. What if the electron is to pass through the transistor gate, without you wanting it to pass? It cannot happen, can it? Well, it very much can. When an electron is presented with a potential barrier, where it doesn’t have enough energy to pass it, we would simply assume that the electron would not cross it right? However, the truth is somewhat different. Let’s assume different cases. First, we assume all electrons can completely cross the barrier, this would simply violate the laws of conservation of energy. Second, we assume that the electrons can’t pass it at all. This is also not true based on several observations made. Now, the third case we suppose is that it can pass the barrier as well as not pass it. Quite like the Schrödinger’s cat, isn’t it? This is explained by superposition where a single particle can exist in multiple states at a given instance in time. Moreover, its presence in a particular position is probabilistic. This phenomenon of electrons passing across a barrier as such is known as Quantum tunneling.
As we miniaturize the transistors, we will be facing this problem more often. Because our classical bits are based on the passing of electrons where active high(1) represents the movement of an electron, and active low(0) represents no movement. Now, this will cause a serious issue for the performance of a system if active high starts occurring without any real passing of current or impulse. Not only the processes, but the data storage also might not be reliable any longer if this problem persists on a large scale. Furthermore, quantum tunneling is a fundamental phenomenon, which means it cannot be eliminated, however, it can be reduced to some extent. A common approach is to increase the width and the height of the barrier. This can be done through regional doping as well as altering the geometry of the system. However, can we take this challenge as an opportunity to find a different approach to more complex calculations and computations? After all, that is what computers are designed for.
The answer to this could be quantum bits or 'qubits'. In qubits, we don’t just deal with two discrete states, but states of superposition, which means it can exist as 0 and 1 at the same time. This also differs from the classical bits because a qubit is defined in terms of the spin of the electrons within its orbital. This can be altered using an EM wave of a specific frequency. However, the sole idea of quantum computing is to reduce the total number of computations to obtain a certain result. Perhaps, the processing speed may still be better in the computers that we use today, apart from certain processes. There are certain algorithms such as Shor’s algorithm and Grover’s algorithm that are designed to work on quantum computers. Despite the possibilities it presents, there have been very few developments although a lot of research activities are currently ongoing.
To sum it up, we as a species have pushed our limits to knowledge and technological advancements every time, we thought we reached our limits. Could Einstein or Planck have guessed what their propositions and theories have done? Would they have even thought their findings would revolutionize the whole ideology of science? Science is just not a subject, it’s a religion that speaks the language of mathematics. Understanding nature, exploring unsee-able phenomena and drawing conclusions have always been the foundation for new theories and solutions. Soon, a reliable solution will be found to this. It’s just a matter of time. Today we explore the underlying issues that come along with the miniaturization of transistors beyond the nanometer mark, tomorrow we may realize that may be miniaturization is not the way ahead.
Source?
ReplyDelete