Networking

NIC (Network Interface Card) - hardware required to connect to a network

Client Server Network - network organisation that consists of several clients and multiple servers (like Print, File, Web) and the clients access and uses the resources available on them.

Network topologies consist of communications channel and nodes (computers)
each has its own advantages and disadvantages

Network Topology Node Failure Channel Failure
Bus no overall effect network failure
Star central node - yes, one node is effected
Ring unless there is a bypass network failure
Mesh/
Fully Connected Mesh no commuincation with node none! several routes that data can take

Peripherals

Puppy Dogs Bite Very Hard

Functions of the Interface (The hardware and Software required to communicate between the Hardware (peripherals) and the processor

Protocol conversion (odd and even parity)
Data Format Conversion (Analogue to digital and visa versa)
Buffering (Temporary Data Storage)
Voltage Conversion (components of motherboard work at lower voltages to peripherals)
Handling of Status Signals (Ready, Out of Ink, Paper Jam)

Solid state storage devices have no moving parts, require less energy are more robust, are physically smaller and more robust.

Buffering - hardware solution, built into peripherals like printer
spooling (also temporary data storage) is the software solution that uses fast backing storage to send data to peripheral at rate it can cope with

Interface speeds are increasing USB 1 - 2 - 3 increased data transfer rates

Backing Storage devices

Smaller Physically, increased capacity, faster data transfer speeds and becoming cheaper per bit storage

The Rest of Computer Systems

Current Trends in Computer Hardware
Processor speeds are increasing
Clock speed is increasing
Increasing memory (Main memory and Backing Storage)
Data Transfer speeds increasing

4 measures to compare computer performance
Application Based Tests
involves benchmarking, getting computer systems to carry out the same tasks and measuring their performance, allows a fair comparison as it measures the performance of the whole computer system

Clock Speed
indicates the speed of the processor (Hertz!) gives an idea of how quickly the computer system executes instructions

MIPS - Millions of Instructions per second
not a fair method of comparison as computer systems have different instruction sets and carry out instrcutions in different ways

FLOPS - Floating Point Operations Per Second
measures how quikcly a computer can perform complex calcualations. these can be compared as the computer systems carry them out in the same method.

Cache memoire!!!!

Faster to access than main memory
stores the most commonly used instructions (don't need to fetch from main memory)
physically closer to main memory

Computer Structure

Read Operation
1. CU Set up MAR with the address to be accessed
2. Control unit activates the Read Line
3. Storage location info. are released onto the Databus
4. Copy is transferred into the MDR

(MAR - Memory Address Register, MDR - Memory Data Register)
The Fetch Execute Cycle (Get it, Dae it!) is an important process in the computer system
a. Fill in the missing steps

1. __________________________________________________
2. Control Unit Activates the read line
3. contents are ________________________________________
4. copy of the information ________________________________
5. instruction is decoded and ______________________________


1. SAME AS READ STEP 1
3. SAME AS READ STEP 3
4. SAME AS READ STEP 4
5. EXECUTED!


Write Operation
1. CU sets up MAR with the address to be accessed
2. CU sets up MDR with data to be written
3. CU activates the write line
4. Data is transferred to mem. location via the Data Bus.

Computer Structure

The Processor's three main parts ... C U A Love U Right!

CU- Control Unit
controls the other parts of the processor, makes sure instructions are carried out in the correct order
ALU - Arithmetic Logic Unit
carries out calculations, makes decisions based on logic (and or not)
R- Registers

temporary storage locations used to store addresses to be accessed and data to be written or read
Two of the three buses on the Computer System

Address Bus - (one way) unidirectional, carry the address to be accessed in memory (goes and points at a mem. location) number of wires tells you how many addresses can be accessed. No effect on computer system performance. (2 ^ width = number of addresses)

Data Bus - (both ways) bidirectional, carry data to and from the processor, main memory and devices. Increasing the number of bits for the data bus IMPROVES systems performance as more data can be carried per clock cycle

Graphics Types

Bitmap

stores the colour (in binary) of every individual pixel

+ edit individual pixels
- fixed resolution (can only print at the resolution its saved at)
- saves the full screen, even if nothing there
- you can't layer object

Vector

stores the image as a list of objects storing their attributes (e.g shape, x and y co-ordinates, line thickness, colour)

+ Resolution Independant
+ can layer objects
+ can edit the individual objects
- can't edit individual pixels

Deano Says ...

Memory Sizes in the correct order

Bit
Byte
Kilobyte
Megabyte
Gigabyte
Terabyte

Parameter passing (modularisation)

ByRef - used when you need to change the value of a variable (physical address)
ByVal - used when the variable needs to not be changed (sends a local copy of variable)

IN:
OUT:

Floating Point Representation (Posse!)

Floating Point rep allows us to represent real numbers

We need two numbers to store it - The Mantissa and The Exponent

The Mantissa represents the actual number to represent the actual number

The Exponent represents the number of places to move the decimal place

Increasing the number of bits for the Mantissa increases the accuracy of the number

Increasing the number of bits for the Exponent increases the range of the number

(MARE!)

Negative Numbers

Signed Bit Notation

has a leading 1 if its a negative number
1 0 0 1 0 0 0 0 = - 16
0 0 0 1 0 0 0 0 = + 16

Its Rubbish!
-Two Representations of zero
1 0 0 0 0 0 0
0 0 0 0 0 0 0

-Lose range because of the signed bit

Two's Complement
*Process
Invert and add 1
Only if its negative (and this has a leading 1)

1 1 1 1 0 1 1 1
0 0 0 0 1 0 0 0 INVERT
0 0 0 0 1 0 0 1 + 1 = - 9

Why do we use the Binary Numbering System?

Binary is a two state numbering system (0 and 1)
Computers are two state machines (On and Off)

Advantages??
* Less Rules of addition (/*-) to program into the processor
* Not affected by voltage degradation
* Easier to store the states in backing storage