Computer Architecture - Transistors

Question

I don’t understand for sure, each transistor can be on or off (0-1), but each transistor then equals to 1 byte or 1 bit?

So if each transistor equals 1 byte so if I had 8 transistors it would be bytes then it would be 64 bits?

Maybe it’s a very simple question to answer, but I’m studying computer architecture seeing how it really works in depth.

Answer 1

Excerpt from the book Santana, Edu - Computer Architecture. Made available by the Federal University of Paraíba: http://producao.virtual.ufpb.br/books/edusantana/introducao-a-arquitetura-de-computadores-livro/livro/

1.6. The Transistor

The transistor is an electronic component created in the 1950s. It is responsible for the electronics revolution in the 1960s. Through it it was possible to develop extremely small digital systems. All features of a computer are internally executed by composing millions of transistors. From logical and arithmetic operations, to the storage of data in memories (except for hard disk, CD, DVD and magnetic tapes), everything is done by transistors.

The first were manufactured on the 10-6 meter micrometer scale). Hence the terms microelectronics and micro-technology. After that, a technological race began to develop transistors that were faster, smaller and cheaper. This revolution lasted until today, but was stronger in the 1980s and 1990s. It was exciting to follow the dispute between Intel and AMD to dominate the personal computer market. Every 6 months a new processor was launched by one of them, taking the competitor’s position as the fastest processor on the market. Few were those consumers who could keep up to date with so many launches.

The basic principle is to use electronics (electric current, resistance and voltage) to represent data and then be able to perform operations with them. The easiest way to do this was to first limit the data to just two types. Zero and one. The binary numbering system is much easier to represent with electronic devices than decimal, for example. The transistor has two states. Either it is loaded, or it is unloaded, as well as a stack. This can easily be mapped to bit 1 (loaded) and bit (0). The revolutionary, unlike a battery, was to enable this state to be changed electronically at any time and very quickly.

So, with 8 transistors in parallel, I can represent, for example, an 8-bit number. I can change their values by changing their loads, and I can read their values coming in if they each have, or not. This is the basic principle of building a memory.

Similarly, it is possible to integrate transistors so that they perform logical and arithmetic operations. The logic ports studied by you in Introduction to Computing are all manufactured using transistors.

The smaller the transistors, the more data can be stored per area. At the same time, smaller transistors store less load. This makes the loading and downloading process faster, which therefore makes data processing and storage much faster as well.

With the evolution of nanoelectronics, transistors are so small that they made it possible to build memories of 1GB (one gigabyte) the size of an adult’s fingernail. To be have an idea, 1 Giga is short for 109, that is, one billion. One byte is 8 bits. So a 1GB memory has at least 8 billion transistors. The processors also became quite fast with the miniaturization of transistors. The current processors work on the frequency of Ghz (Giga Hertz), that is, on the average of billions of cycles per second (different from operations per second). This is too fast!

Figure 1.3. Structure of a MOSFET-type transistor

In Figure 1.3, the previous "Structure of a MOSFET type transistor" is presented the structure of a MOSFET transistor. This transistor is the most used to build digital electronic systems, such as computers. The name comes from the abbreviation "Metal-Oxide Semiconductor Field-Effect Transistor". Let’s see what each of these words means, and this will help us to know a little more about MOSFET and its relevance. The term MOS ("Metal-Oxide Semiconductor") comes from the materials used to compose a MOSFET, which are mainly metallic oxide and semiconductor.

Semiconductors are materials that have properties that neither allow them to classify as conductor nor as insulating. In some conditions he acts as an insulator, and in others as a conductor. The most used semiconductor in transistors is silicon (symbol Si in the Periodic Table). In ambient conditions, silicon acts as an insulator, but if mixed with other materials, it can become a conductor up to the desired intensity.

Silicon became so important that it changed an entire region of California in the United States in the 1950s, making it one of the the most promising in the world to date. This region has sheltered and shelters the most important companies in the field of computer design, such as Intel, AMD, Dell, IBM and Apple, and after software would go run on these computers like Microsoft, Oracle and Google. This region is called Silicon Valley.

In the transistor of Figure 1.3, "Structure of a MOSFET type transistor" light gray represents a silicon crystal that has been doped with negative charges. The dark gray represents the part that was doped with positive charges.

In the normal situation (see Figure 1.4, "Opening and closing the gate of the MOSFET type transistor") an electric current applied to the Drain can travel through the narrow negative channel and follow to the Source. In that condition we say the transistor is active. However, if a negative voltage is applied to the Gate, the positive charges of the p region will be attracted closer to the Gate, and this will close the channel through which the electric current passed. In that case, we say the transistor is inactive.

Figure 1.4. Opening and closing of the MOSFET type transistor gate

Why do we care? When the transistor is active, it can be seen with the value 1, and when inactive, it can be seen with the value 0. Thus, we have the smallest memory possible to build. When we want it to keep the value 1, just turn off the voltage of the Gate and apply a current in the Drain. When we want it to store the value 0, we need to apply a chain to the Port and close the channel. So an 8-billion-bit memory can be built on 8 billion transistors like these.

Now we know the first aspect that makes transistors essential for the understanding of the computer. They are used to build memories. Memories made from transistors are also called Solid State Memories. But there are others, not so efficient and miniaturized, such as optical and magnetic memories. The important thing to realize is that the smaller we can build these transistors, the better. The channel opening and closing process is not instantaneous. It takes a very short time, but when you add up the times of all the billions of transistors, it becomes relevant. The smaller it is, the narrower the channel is, and therefore the faster it turns on and off, in the same way, the shorter the distance between the Drain and the Source, also taking less time for the electrons to leave the Drain towards the source. This will all make the memory faster. Small transistors also enable more data to be stored per area. That is why today huge storage capacities are available in devices as reduced as are the examples of pen-drives and memory cards.

Transistors are also used to perform logical and arithmetic operations. The load taken from one transistor can serve to feed one another and that, if combined correctly, can perform the basic logic operations, E, OR, NO and arithmetic, addition, subtraction, division and multiplication. With this, transistors can not only be used to store data, but how to perform logical and arithmetic operations on that data. This is fantastic and has revolutionized the whole world. Not only in computer science, but also in all areas of knowledge. What would humanity be today without the computer? Without the cell phone? Without the satellites?