APPENDIX 8: ABBREVIATIONS AND GLOSSARY
|ALU||Arithmetic logic unit
|BCS||British Computer Society
|CITRIS||Centre for IT Research in the Interest of Society, University of California
|CPU||Central processor unit
|CRT||Cathode ray tube|
|DTI||Department of Trade and Industry
|DSP||Digital Signal Processor
|EPSRC||Engineering and Physical Sciences Research Council
|GDP||Gross Domestic Product
|GSM||Global System for Mobile communications
|HEFCE||Higher Education Funding Council for England
|ICT||Information and communication technology
|IEE||Institution of Electrical Engineering
|IMEC||Inter-University Microelectronics Centre, Belgium
|IRC||Interdisciplinary Research Collaboration
|ISLI||Institute for System Level Integration, Livingston
|IT&CS||Information Technology and Computer Science
|ITRS||International Technology Roadmap for Semiconductors
|LCD||Liquid crystal display
|NHS||National Health Service
|NPL||National Physical Laboratory
|OECD||Organisation for Economic Co-operation and Development
|OMI||Open Microprocessor Systems Initiative
|p||page number in the accompanying volume of evidence
|PDA||Personal digital assistant
|PPARC||Particle Physics and Astronomy Research Council
|Q||question number in the accompanying volume of evidence
|R&D||Research and development
|RAE||Research Assessment Exercise
|RISC||Reduced instruction set computing
|SETI||The Search for Extraterrestrial Intelligence project
|SLI||System level integration
|US||United States of America
Very large and very small numbers
1. Modern computer technology involves both very large and
very small numbers. To avoid long names or large numbers of decimal
places, these are normally indicated by use of standard prefixes.
For example, a millionth of a second is referred to as a microsecond
and a millionth of a metre is a micrometre (also sometimes referred
to as a micron).
2. These prefixes cover the very large to the very small as
in the following table. The difference between each step is a
factor of a thousand.
||thousand million millionth
3. It is obviously easier to say "3 gigahertz" (a
measure of computer speed) than "three thousand million hertz".
For calculations and some other purposes, however, it is necessary
to write down the numbers. If written in full, 3 gigahertz would
be 3,000,000,000 hertz. However, the conventional way of writing
such a large number is 3 x 109, being three multiplied
by 10 to the power 9 (10 multiplied by itself nine times) or a
4. Such power notation is also used to simplify very small
numbers. A minus power represents one divided by the number that
would be generated by that power if positive. For example, 10-9
(referred to as "10 to the minus 9") is one divided
by 109 or one thousand millionth. 5 nanometres (being
five thousand millionths of a metre or 0.000000005 metres) may
thus be written as 5 x 10-9 metres.
5. Individual elements on a modern computer chip are best
measured in nanometres (nm), of which there are one thousand million
to the metre, the standard metric unit of length. (The now less-used
Ångström unit is one tenth of a nanometre.)
6. To illustrate such very small distances, a ream (500 sheets)
of standard paper is 50 millimetres thick. 10 sheets are thus
a millimetre and a single sheet is 100 micrometres thick. A nanometre
is one thousandth of a micrometre. If it were possible to make
paper only one nanometre thick, it would take one hundred thousand
sheets (200 reams) to equal the thickness of one sheet of ordinary
7. Computer chips are manufactured in layers, some of which
are only 1.5 nm thick. Atoms in a silicon crystal are about a
fifth of a nanometre apart, so such layers are only 7 or 8 atoms
thick. Critical horizontal measurements can be as small as 5 nm
or 25 atoms.
8. A computer's actions are synchronised by high-frequency
pulses from its clock. The standard measurement of frequency is
the hertz, being one event per second. Modern PCs operate at clock
rates in the gigahertz range, that is at over a thousand million
(109) cycles per second. To say that a computer works
at 1 gigahertz is exactly the same as saying that it completes
each event in 1 nanosecond (10-9 seconds).
9. Each step of a computer's operation is one of those events.
However, delivering a program instruction (for example, retrieving
data, processing it and storing the result) involves several steps.
A measure of a computer's operating speed is the millions of instructions
per second (MIPS) with which it can deal.
10. Calculations involving fractions are more complicated
than dealing with integers. For some applications, users need
to know a computer's speed in handling floating point operations
per second (FLOPS).
11. Early computers completed each operation before turning
to the next so their speed in MIPS was always less than their
clock rate. Modern machines not only handle a number of instructions
in parallel but also begin the next batch of instructions before
the previous batch is finished. Their speed in MIPS or FLOPS can
thus exceed their clock rates.
12. The basic unit of computer data is a binary bit
which will have the value of either one or zero. (What a bit represents
is dictated by its context.)
13. A block of 8 bits is called a byte. Modern computers handle
"words" of 32 or 64 bits simultaneously.
14. Computers' data storage capacity is now normally measured
in megabytes and gigabytes and even in terabytes.