Lecture 1 - 4

• Computer Architecture


  Attributes of the system that are visible to the software programmer and have a direct impact on the logical execution of a program, like the number of bits used to represent various data types, the instruction set of the computer, technique for addressing memory, method used for input, output, etc.
• Main categories considered in computer architecture design

  • System Design
  • Instruction Set Architecture
  • Micro Architecture
• System Design


  Hardware components used for building the system
• Instruction Set Architecture


  Includes all the instructions provided to the computer system
• Micro Architecture


  Gives minute detail about storage elements
• Computer Organization


  The way the various components of a computer system are interconnected and work together to perform tasks and execute programs
• 5 basic components of a computer

  • Input Unit
  • Output Unit
  • Memory Unit
  • Control Unit
  • Arithmetical and Logical Unit
• Input Unit


  Where commands are given to the device
• Output Unit


  Where the result of a task is displayed
• Memory Unit


  Stores the entered information and transmits it to other parts of the CPU
• Control Unit


  Manages the entire functioning of the computer device
• Arithmetical and Logical Unit


  Performs mathematical calculations and comparison/decision-making actions
• Instruction Set Architecture (ISA)


  Part of the abstract model of a computer that defines how the CPU is controlled by the software, acting as an interface between hardware and software
• RISC Processor


  • Microprocessor architecture with a simple and highly customized set of instructions
  • Each instruction cycle requires only one clock cycle, with fetch, decode and execute stages
  • Combines complex instructions into simpler ones
  • RISC Chips - Requires fewer transistors, making it cheaper to design and reducing execution time
• CISC Processor


  Has a large collection of complex instructions from simple to specialized, taking longer to execute
• Types of Memory Hierarchy

  • External Memory/Secondary Memory
  • Internal Memory/Primary Memory
• External Memory/Secondary Memory

  • Hard Disk Drive (HDD)
  • Solid State Drive (SSD)
  • USB Flash Drive
  • External Hard Drive
  • Memory Card
  • NAS or File Servers
  • Cloud Storage
• Internal Memory/Primary Memory

  • Random Access Memory (RAM)
  • Cache Memory
  • Registers
  • Memory Controllers
  • On-Chip Memory (L1,L2, L3 Cache)
  • Non-volatile Memory (ROM)
• Registers


  Tiny, lightning-fast memory units inside the CPU that store data temporarily during processing
• Types of Registers

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


  Super-fast memory that stores frequently used data from main memory, located close to the CPU to reduce access time
• Cache Mapping Techniques

  • Direct Mapping
  • Associative Mapping
  • Set-Associative Mapping
• RAM


  crucial part of modern computers that stores programs and information for the processor to use, working closely with the CPU and I/O devices
• Types of Main Memory

  • Static RAM
  • Dynamic RAM
• Secondary Storage

  Non-volatile memory with larger storage capacity than main memory, used to store data and instructions not currently being used by the CPU
• Design Principles and Trade-offs in Memory Hierarchy


  • Principle of Locality
  • Cost-Performance Trade-off
  • Caching
  • Block Size
  • Replacement Policy
  • Write Policy
• Pipelining


  Technique that allows a processor to execute multiple instructions at the same time by breaking them down into smaller steps
• How Pipelining Works

  1. Instruction Fetch

  2. Instruction Decode

  3. Instruction Execute

  4. Memory Access

  5. Write Back
• Types of Pipelines

  • Instruction pipeline
  • Arithmetic pipeline
• Benefits of Pipelining

  • Improves processor efficiency by reducing idle time between instructions
  • Overlaps execution of different instructions in different stages
• Challenges of Pipelining

  • Dependency problem - when an instruction depends on the result of a previous unfinished instruction
  • Branch problem - when the processor has to make a decision based on a condition, leading to mispredictions and stalls