SUMMARY OF LECTURE 5
THE COURSEKIT ARTICLES
From The Article, "From Sand To Silicon", you should familiarize
yourselves with the steps in the manufacture of an integrated circuit/microchip.
These steps are: designing the chip; making a silicon wafer; creating a layer of
silicon dioxide which serves as an insulator i.e. it does not conduct electricity;
masking, etching, adding layers, doping, and interconnections. Generally
speaking, if you learn the first paragraph of each step, that is sufficient. I do not
want you to delve into the science portion. For example, I want you to
understand that "masking" is a process in which ultra-violet light is shone through
a circuit diagram called a "mask" onto a silicone wafer, so as to create an imprint
of the circuit on the wafer. Again, I want you to understand that etching is like
carving out the circuit diagram that has been imprinted onto the silicone wafer
and how this is done ( i.e. what is used to do this) using acids and reactive
gases. Adding layers is somewhat self explanatory and after this comes
"Doping", (not to be confused with "Dopey", one of the dwarves in the story Snow
White). You should understand that this process uses chemical impurities (such
as arsenic or boron) to alter or change the way in which those parts of the silicon
wafer which are doped conduct electricity. Another way of thinking of this
process is that the result of performing this step over and over again is to create
positive and negative charged areas on the silicon wafer thereby creating the
transistors. Finally, interconnections, refers to the process of creating the wires
that connect the chip's transistors together, since leaving the transistors
unconnected would result in the chip not working.
From the article, "Technology and Economics in the Semi Conductor
Industry", I expect you to understand first, "Moore's Law"; secondly, that there
are Technical and Economic Barriers to the continued advancement of
transistorized microchips/computing devices and what some of these are; and
finally, from the sub-Article, "How Much bang For the Buck", I expect you to gain
an understanding of why the Semi-Conductor Industry is different from other
industries and the reasoning behind the graph that is presented respecting Intel's
As to "Quantum-Mechanical Computers", all you need to do is to you read
the article and try to understand how the atomic chip works. In this regard, the
captions (the explanations) of the diagrams on the first and second pages of the
article are quite helpful, although I told the class that I would explain this more by
way of simple diagrams which will be provided in a later lecture. 2
ARTICLE 1: THE TRANSISTOR
This article explains how a transisitor works and its advantages over its
predecessor, the VacuumTube. You need only know what follows.
The transistor is a method of controlling electrons in a solid crystal
instead of in a vacuum. It derives its name because it transfers an electrical
signal across a resistor.
ADVANTAGES OVER VACUUM TUBE:
1. It reduces the complicated delicate tube to a simple rig consisting of a couple of
fine wires ("cats whiskers") and a small crystal. Not vacuum tube is needed.
2. It does not have to heat up like a vacuum tube does, so that it works instantly.
3. It operates on a tiny amount of power (1/10 of a flashlight bulb).
4. It can be made extremely small, and by doing so allows the reduction of
electronic equipment to smaller scales.
5. It is long lived and sturdy being less costly than the vacuum tube. Thus, it
simplifies the manufacture and maintenance of electronic equipment.
6. Its instant response makes it useful in computers.
7. The small power requirements of a transistor makes it possible to use
for prolonged periods e.g. for alarm systems, telephone repeaters etc....
BASIS OF TRANSISTOR OPERATION
The transistor allows control of the number of current carrying electrons in
semi conductors (crystals). Impurities in a crystal semi conductor free some of
the electrons that would otherwise be used to link atoms allow it to conduct
electricity. The electrons can be made to act as either conductors or insulators.
In fact research suggests that the surface of a semi conductor is a better carrier
of electrons than its interior. 3
ARTICLE 2: COMPUTERS FROM TRANSISTORS:
THE BASIS OF TRANSISTORS AND COMPUTERS
Digital computers operate by manipulating statements made in binary code
which consists of ones and zeros. A field effect transistor is operated so that it
is switched only "on" or "off". The device represents exactly one binary unit of
information: a bit.
In a large system, input signals control transistors that switch signal voltages
onto output wires. The wires carry the signals to other switches that produce
outputs, which are again sent on to another stage. The connections determine
its function. They control the way that the inputs are transformed to become
outputs, such as a word in a document or an entry on a spreadsheet.
ARTICLE 3: FROM SAND TO SILICON: THE MANUFACTURE OF AN
This article sets out the steps in the manufacture of an integrated circuit or
what is now called a microchip or chip. From this article you should understand
the following steps.
1. Chip Design, which is self explanatory.
2. The Silicon Crystal: This is the step in which thin slices or wafers of silicon
are created for use in the manufacture of an integrated circuit.
3. The First Layers: Here a layer of silicon dioxide (which does not conduct
electricity and in consequence acts as an insulator, is placed on the silicon wafer.
This is followed by the addition of a coating of a viscous polymer liquid called
photoresist, which dissolves when exposed to ultra violet light.
4. Masking: This is a process, using lithographic equipment, an ultra violet light
source, and a thin film or mask on which is the design layout of the circuit
diagram of the chip. In this process, ultra violet light is shown through a mask
onto the photoresist on the silicon wafer planting or fixing the circuit pattern or
diagram onto the wafer. 4
5. Etching: Acids and reactive gases are used to carve out the circuit diagram
which masking placed on the wafer.
6. Adding Layers: In this process, masking and etching is repeated resulting in
the deposit of additional materials on the chip, increasing its depth.
7. Doping: Here chemical impurities are added using born and arsenic to alter
the way the silicon on each doped area conducts electricity. To do this the wafer
is showered with ions of boron and arsenic creating positively and negatively
charges areas on the wafer. These positively and negatively charged areas will
become the transistors.
8. Interconnections: Here metal such as aluminium and, more recently
copper, is deposited into grooves in the wafer created by further masking and
etching, creating a form of wiring that connects the transistors together to
complete the circuit.
In the next lecture, the above steps will be compared to the steps on page 37
of the Coursekit, in the article, “The First Nanochips”. However, this is a good
time to summarize them and so, generally, steps 1 and 2 on page 37 are
equivalent to Chip Design and The Silicon Crystal above. Step 3 is equivalent to
Masking. Steps 4 and 5 are the etching process, while step 6 is the doping
process. Step 7 is the same as Interconnections.
ARTICLE 3: TECHNOLOGY AND ECONOMICS IN THE SEMI-CONDUCTOR
For one who has little to no background in science, understanding the
semi-conductor industry and what is happening within it would seem to be a
somewhat unnecessary endeavour. Technology apathy reigns supreme - people
are simply accustomed to acquiring and using the new technology offered them,
without any concern about its development or its consequences, a philosophy
which, unfortunately, is reinforced by the implicit assumption that the cost of
acquiring the new technology will always be within their limits or reach. This idea
of new technology being within everyone's reach - being affordable - is a direct
consequence of the economics of a transistor-based technology. To put it
another way, were can all thank our lucky stars that the expense involved in
placing more and more transistors on a chip is not significant, otherwise new
technology would be beyond the reach of the majority of us.
The principle of affordability of new technology is a direct consequence of
a proposition put forward by, Gordon Moore, the inventor of the chip 5
manufacturing process and whose company, Intel Corporation, is known to all of
you. Back in the 1960's, Mr. Moore observed that each year the number of
transistors that could be put on a chip was doubling, but without any significant
increase in cost. A simple restatement of this is that every year the number of
transistors that can be put on a chip doubles, while the expense per
transistor is halved (cut in half) with each doubling. Moore speculated that
this would continue indefinitely and to date it has, although the time period seems
to have increased to from a year to first eighteen (18) and now 24 months. For
this reason, Moore's observation has come to be regarded as a "law" of
transistorized chip development. Increasing the number of transistors on a chip,
increases the chip's ability for memory or data storage; permits a greater number
of functions to be performed by the chip, and allows transistors which are packed
more closely together to more quickly interact with each other. However, even
the best predictions do not last forever and, given the shrinking size of the chip,
there is a real concern that Moore's law will not continue beyond 2010 for both
technical and economic reasons.
THE TECHNICAL BARRIERS/REASONS: