System Architecture

• Purpose of the CPU

  To fetch, decode and execute instructions
• CPU


  The brain of the computer, takes an input, processes data and produces an output
• Fetch-Execute Cycle


  The cycle that the central processing unit (CPU) runs through billions of times per second to make a computer work
• Fetch-Execute Cycle


  1. Fetch stage

  2. Decode stage

  3. Execute stage
• Fetch stage


  The next instruction or data must be fetched from the computer's memory (RAM)
• Decode stage


  The CPU needs to work out what is required from the instruction
• Execute stage


  The CPU will carry out the instruction that was fetched, such as performing a calculation, storing a result or data back in main memory, or going to main memory to fetch data from a different location
• Input


  • Clicking a button on the gamepad
• Processing input


  The CPU inside the console follows a set of instructions to carry out the task
• Output


  • The player moving on screen
• The important things to remember are: An instruction or data is fetched from memory, the instruction is decoded, the instruction is executed, the cycle repeats billions of times per second
• CPU


  Central Processing Unit
• Components of the CPU

  • Arithmetic Logic Unit (ALU)
  • Control Unit (CU)
  • Cache
  • Registers
• Function of the Arithmetic Logic Unit (ALU)

  1. Performs arithmetic operations

  2. Performs logical decisions
• Function of the Control Unit (CU)

  1. Coordinates how data moves around the CPU by sending a signal to control the movement of the data

  2. Decodes the instructions fetched from memory
• Cache


  Very small, very fast memory located in the CPU which is used to provide quick access to frequently used instructions and data
• More cache there is

  More data can be stored, which speeds up the performance of the CPU
• Cache


  Prevents the CPU from having to repeatedly fetch frequently used instructions from RAM
• Registers


  Extremely small, extremely fast memory located in the CPU, each with its very own specific purpose
• More details about the individual registers can be found here
• Von Neumann Architecture


  A design of the CPU which was proposed by Mathematician John Von Neumann in the 1940s, which most general-purpose computers are built upon
• Von Neumann Architecture


  • It outlines how the computer memory, input / output devices and processor all work together
• Main registers in Von Neumann Architecture

  • Program Counter (PC)
  • Memory Address Register (MAR)
  • Memory Data Register (MDR)
  • Accumulator (ACC)
• Program Counter (PC)


  • Holds the memory address of the next instructions to be executed
  • Increments by 1 as the fetch-decode-execute cycle runs
• Memory Address Register (MAR)


  Holds the memory address of where data or instructions are to be fetched from
• Memory Data Register (MDR)


  Stores the data or instruction which has been fetched from memory
• Accumulator (ACC)


  Stores the results of any calculations that have taken place in the Arithmetic Logic Unit (ALU)
• Common characteristics of the CPU

  • Clock Speed
  • Cache Size
  • Number of Cores
• Each of these characteristics has a significant impact on the performance of the CPU
• Clock Speed


  The clock speed is measured in Hertz (Hz) and measures the number of fetch-decode-execute cycles that can take place in 1 second
• The faster the clock speed


  The more instructions can be fetched and executed per second
• Modern computers have a clock speed in Gigahertz (GHz)


  Meaning billion
• A clock speed of 3.5GHz can perform up to 3.5 billion instructions per second
• Cache


  Very small, very fast memory on or close to the CPU, used as temporary storage to provide quick access to a copy of frequently used instructions and data
• The larger the cache size


  The more frequently used instructions or data can be stored
• Larger cache size


  Results in the CPU having to complete fewer fetch cycles from memory (RAM), speeding up the performance
• Cache has a significantly faster read/write speed than RAM

  Making it much quicker to retrieve instructions from there instead of from memory (RAM)
• Core


  A core works like it is its own CPU, with its own Control Unit (CU), Arithmetic Logic Unit (ALU), Accumulator (ACC), and Registers
• Multiple core processors


  Have multiple separate processing units that can fetch, decode and execute instructions at the same time
• Multi-core processors


  Can run more powerful programs with greater ease