• It contains an Arithmetic and Logic Unit (or ALU) to perform all addition, subtraction, logic operations, etc.
• It also contains several registers, these temporarily store very small pieces of information, and give the ALU very fast access to that data
• Finally, it contains all the control logic to coordinate actions between all elements of the CPU as well as all the other computer hardware components

• Main memory is the set of storage locations immediately accessible by the CPU
• It is not as fast as the registers, but is still quite fast, is much cheaper, and can hold much more information
• Main memory is divided into two main types:
  • RAM - which can be altered by the CPU, and
  • ROM - which can only be read by the CPU
  All RAM contents are wiped out when the computer is turned off, while the ROM contents are permanently set at the factory.
• Sometimes an extra ``layer'' of memory is added, a cache, which is faster but more expensive than regular memory.
  Cache is used to store data which is used very frequently.

In secondary storage data is stored in files (e.g. information in a database, folders with your electronic mail, word processing documents you want to keep electronic copies of, etc.)

• Secondary storage includes disk drives, DVD drives, CD drives, tape drives, flash/thumbnail drives, etc
• It is non-volatile storage (i.e. the contents are not lost when the computer is shut down)
• It is generally much slower than main memory, but is also less expensive (per bit of data stored) and can hold much more information.

• I/O devices include printer, mouse, keyboard, monitor, touch-screen, light pen, speakers, microphones, temperature sensors, etc
• Usually there is a small controller that acts like a go-between for the CPU and the actual I/O device - simplifying modifications to the system

The bus is used to transfer data between the CPU, I/O devices, secondary storage devices, and memory.

• Networks link computers and devices within a room or around the world, and may be wired or wireless
• Local Area Networks (LANs) - use cables to connect machines within a small local area (say < 1km radius)
• Wide Area Networks (WANs) - use telephone systems, satellites, and microwave relays to connect machines across much greater distances
• Bridges and gateways connect different networks together

• Once a programmer has decided how to solve a problem, they need to write the solution out in a valid programming language (e.g. C++)
• Source code is the program created in a ``human-readable'' format, e.g. when we write code using C++
• This ``human-readable'' format cannot be directly run by the computer, so when you go to run the program a compiler (such as gcc or g++) must first be used to translate the source code into executable code
• The compiler translates the source code into another format, one which the computer can run or execute
• Source code is generally portable (can be used on different machines) whereas executables rely on machine-specific information

• Executable programs (as well as source code) are stored on secondary storage (usually disk drives)
• To execute a program it must be copied into main memory, and space set aside for variables, subroutine calls etc
• Every individual instruction, and every piece of data, exists at some unique location (or address) in memory
• To execute a program, the computer keeps track of the memory address of the next instruction it must execute
• Execution then consists of repeatedly:
  • Fetching the next instruction to be executed
  • Decoding the instruction (deciding what must be done)
  • Executing the instruction (and deciding which instruction must be executed next)

• When a program is running, a copy of it is stored in memory
• The CPU uses a register (the program counter) to keep track of which program instruction is going to be executed next
• Using the communications bus and the value in the program counter, the CPU asks memory to look up (fetch) the next instruction
• Long instructions may take several fetch stages
• The instruction is stored somewhere within the CPU (possibly in an instruction register) and the program counter is adjusted to `point at' the next instruction
• The CPU decodes the instruction, to determine what action needs to take place
• Finally, the CPU carries out, or executes the instruction.
• The cycle is now complete, and can start again for the next instruction.

1. Developing a plan or algorithm for solving the problem
   • requires a good understanding of the problem from the user's perspective
   • requires an understanding of the general solution techniques and data storage structures which are likely to be effective
   • requires a clear description of each phase of the solution in a logical and correct fashion
2. Expressing the problem in a particular programming language
   • must select a language supported by the development and target systems
   • the language should be appropriate for the kind of problem being addressed
   • the programmer needs fluency in the language

• Systems software
  • used for managing the hardware and controlling user access to it
  • aims at making the system more efficient, easier to use, more reliable, etc.
• Applications software
  • software actually initiated by the user to solve a problem or achieve some end
  • includes word processing, databases, email, web browsers, games, etc

• Machine language
  • binary sequences (0's and 1's) which represent an instruction in the control logic of a specific CPU type
• Assembly language
  • mneumonics corresponding to machine language (e.g. JSR DECOUT instead of 0100111010000000)
• High level language
  • instructions corresponding to a common kind of solution step, e.g. write(x) to print out the value of x

• are meant to correspond (roughly) to the way humans describe solutions
• are more abstract than low level languages, and therefore are portable between machines
• compilers translate high level language instructions into machine code for later execution
• often a single HLL instruction will be represented by several machine code instructions
• are often tailored to specific kinds of problems:
  • FORTRAN - scientific calculation
  • COBOL - report generation, handling business data
  • C, C++ - general purpose problem solving languages

DO  99, X=5,10,1
    PRINT X
99 CONTINUE

for x := 5 to 10 do
begin
  write(x);
end;

for (x=5; x<=10; x=x+1)
{
    printf("%d", x);
}

for (x=5; x<=10; x=x+1)
{
    printf("%d", x);
}

      MOVE.L  #4,D2
LOOP: MOVE.L  D2,D0
      ADDI.W  #6,D0
      JSR     DECOUT
      DBRA    D2,LOOP

00100000101111000000000000000100
0010000000000010
00000110010000000000000000000110
0100111010000000
01010001110010101111111111110100

• The C programming language was developed by Kernighan and Ritchie at Bell Labs in the early 1970's
• It is a powerful general-purpose language, developed to work closely with the Unix operating system
• C++ is a descendant of the C programming language (really it is C with a bunch of extra bells and whistles)
• C++ is an object-oriented language - that is, it incorporates many special features to encapsulate and manipulate user-defined kinds of data
• Most of the OO features of C++ will be discussed in the CSCI 161