Chapter 10 - Computer Systems

Página 1

The central processing unit (CPU)

Inside a processor there are billions of transistors

Carry out instructions given to it

Runs to the timings of a clock signal

The speed of the signal/clock speed is measured in hertz

Unit______________________Pulses per second

1 hertz___________________1

1 kilohertz_______________1,000

1 megahertz_____________1,000,000

1 gigahertz______________1,000,000,000

Processors work by continually fetching instructions from memory, decoding them and executing them. This is known as the fetch-decode-execute cycle

Inside the processor

Registers

Memory locations within the processor itself

Work at extremely fast speeds (so that it can be used by the processor without causing a bottleneck)

Program Counter (PC)

Keeps track of the memory location of the machine code being executed

Gets incremented to point to the next instruction

Changed by instructions that alter the flow of control

-Branch if zero (BRZ)

-Branch always (BRA)

-Branch if positive (BRP)

Memory data register (MDR)

Stores the data that has been fetched from or stored in memory.

Memory Address register (MAR)

Stores the address of the data or instructions that are to be fetched from or sent to.

Current instruction register (CIR)

Stores the most recently fetched instruction, waiting to be decoded and executed.

Accumulator (ACC)

Stores the result of calculations made by the ALU.

General purpose registers

Used temporarily to store data being used rather than sending data to and from comparatively much slower memory.

Buses

Are the communication channels through which data can be sent around the computer

USB(Universal serial bus) - Used to transfer data between the computer and external devices

When looking at the fetch-decode-execute cycle, there are three buses inside the computer that must be considered:

The data bus - carries data between the processor and memory

The control bus - sends control signals from the control unit

The address bus - carries the address of the memory location being read from or written to

Arithmetic logic unit (ALU)

Carries out calculations and logical decisions. The results of its calculations are stored in the accumulator.

Control unit (CU)

Sends out signals to co-ordinate how the processors works.

Controls how data moves around parts of the CPU and how it moves between the CPU & memory.

Instructions are decoded in the control unit.

Página 2

Input, output, storage and memory

Input devices allow data to be entered to the the computer. Examples include:

Keyboards

Mice

Microphones

Scanners

Joysticks

Output devices allow information to be retrieved from a computer and displayed

printers

speakers

monitors

actuators

There are 3 categories of storage devices:

Magnetic

Advantages

Disadvantages

Uses a magnetisable material. Patterns of magnetisation are used to represent binary sequences.

Examples: Hard disk drives, magnetic tape

Magnetic tape

tape used in recording sound, pictures, or computer data.

The oxide particles themselves do not move during recording. Rather their magnetic domains are reoriented by the magnetic field from the tape head.Tape heads for stereo cassette tapes must record on two tracks simultaneously.

Magnetic tape is a medium for magnetic recording, made of a thin, magnetizable coating on a long, narrow strip of plastic film.

Advanatges

the cheapest form of storage per megabyte of storage

Disadvanatges

Serial access so can be quite slow to access data

Magnetic disk

Flash

A kind of memory that retains data in the absence of a power supply.

Has no moving parts.

Uses a special type of ROM that can be overwritten. Used in USB memory sticks and camera memory cards

Advantages

Consumes less power than other types of media.

Well suited to portable devices

Disadvantage

More expensive than magnetic or optical

Optical

A computer memory that uses optical techniques which generally involve an addressable laser beam, a storage medium which responds to the beam for writing and sometimes for erasing, and a detector which reacts to the altered character of the medium when it uses the beam to read out stored data.

Works by using a laser and look at its reflection, determining where there are pits on a surface representing 1s and 0s.

A DVD can contain more data than a CD because the pits are closer together.

Advantages

Cheap to distribute

Fairly resiliant

Disadvantages

Cant hold as much data as other forms of storage, e.g. magnetic

Página 3

RAM & ROM

RAM

Volatile - When the power to the computer is lost, RAM loses its contents

When the programs and data being run are temporarily used

Much quicker access to RAM than any other storage device

ROM

Can be read from but not written to

Used for storing the program to boot up a computer

Non-volatile - keeps its memory when the computers power is lost

Página 4

Harvard architecture

Data and instructions are stored in separate memory units with separate buses

The next instruction can bead read from the instruction memory while data is being read or written to memory

Used by RISC processors

Página 5

Von Neumann architecture

Instructions and data are stored in memory together

Página 6

Parallel processing

(SIMD) Single instruction multiple data - The same operation is carried on multiple pieces of data, at one time.

(MIMD) Multiple instructions multiple data - Different instructions are carried out concurrently on different pieces of data

Takes place on a much larger scale on supercomputers

Supercomputers are massive parallel machines. Top supercomputers contain tens of thousands of multiple processors. They are really expensive to buy and run.

Distributed computing is an alternative approach to parallel computing.

Each computer across a network takes on a part of a problem

Not all problems can be carried out faster using multiple codes. E.g. the fibonacci sequence. Each term depends on the previous so it cant be carried out faster with multiple cores

(RISC) - Reduced instruction set computing

The number of instructions is streamlined

Tend to have fewer addressing modes and more general purpose registers than non RISC processors

All RISC systems should execute all their instructions in roughly the same cycle

RISC systems can use pipelining

Compiler has to do more work

Every operation takes place in just one clock cycle

Longer code requiring more RAM to execute

Further away from high level languages

Simple hardware

(CISC) - Complex instruction set computing

Less work for the compiler to carry out

More complex hardware

Some instructions take multiple clock cycles to complete

Closer to high level language

Smaller code requiring less RAM to execute

Página 7

1st day back

test on chapter 9 & 10