Computer Architecture Quiz

Video Chapters

• Chapter 1: Introduction covers Boolean Algebra, computer types, functional units, buses, bus architecture, types of buses, bus arbitration, registers, organization, and addressing modes.
• Chapter 2: Arithmetic and Logic Unit (ALU) details look-ahead carry adders, multiplication (Booth's algorithm and array multipliers), division, and logic operations, floating-point arithmetic, and ALU design. It also covers the IEEE standard for floating-point numbers.
• Chapter 3: Control Unit explains instruction types, formats, instruction cycles (fetch and execute), micro-operations, program control (e.g., reduced instruction set computer), and control types (hardwired and microprogrammed).
• Chapter 4: Memory explores basic concepts, hierarchy, semiconductor RAM memories (2D and 2 1/2D), ROM memories, cache memories (concepts and design), auxiliary memories (magnetic disk, magnetic tape, optical disks), and virtual memory concepts and implementation.
• Chapter 5: Input/Output (I/O) discusses peripheral devices, I/O interfaces, I/O ports, interrupt hardware, interrupt types, exceptions, modes of data transfer (programmed I/O, interrupt initiated I/O, and direct memory access), and I/O channels and processors. It also covers communication interfaces (synchronous and asynchronous).
• Chapter 6: Pipelining details the differences between uniprocessing and multiprocessing, pipelining, speedup, structural hazards, control hazards, data hazards, and operand forwarding.

Boolean Algebra

• The signal in most present-day electronic digital systems uses two discrete values, making it binary.
• George Boole introduced the concept of a binary number system in logic studies, leading to Boolean algebra.
• Boolean algebra is a branch of algebra where variables represent truth values (true/false, typically denoted 1 and 0).
• Conjunction (AND) is denoted as A.
• Disjunction (OR) is denoted as V.
• Negation (NOT) is denoted as ¬.

Turing Machine

• The Church-Turing thesis states that any algorithmic procedure a human or computer can perform can also be carried out by a Turing machine.
• The Turing machine is a theoretical model of a computer.

Digital System

• In the 1930s, Claude Shannon applied Boolean algebra to switching circuits, introducing switching algebra for circuit analysis and design.
• This logic, organized and systematized by Shannon, is the basis for many digital hardware designs.
• Boolean variables can take different values at various times.
• They can represent input, output, or intermediate signals in circuits.
• Boolean constants are 0 and 1.

Historically Types of Computers

• Fixed Program Computers/Dedicated Device/Embedded System: Their function is specific and cannot be reprogrammed (e.g., calculators, washing machines.)
• Stored Program Computers/General Purpose computer/von Neumann architecture: Programmed to carry out many tasks with stored applications.

Machine/CPU Architecture

• A CPU has three main components: control unit (generates control signals), registers (temporary storage), and ALU (performs arithmetic and logic operations).
• Registers include program counter (holding address), instruction register, and base registers.
• Examples of ALU operations are addition, subtraction, comparisons, and bit shifting.

Main Memory Elements

• The main elements needed for a computer system include memory to store instructions, and an ALU to perform operations.
• Registers are used as a fast memory.

Bus Arbitration

• Bus arbitration is a method to resolve simultaneous requests for access to a shared bus.
• Without arbitration, data corruption and system malfunctions can occur.
• Methods include daisy chaining (simple and cheaper) and polling.

Addressing Modes

• Addressing modes specify different ways to refer to operands in an instruction.
• Effective address is the final address used to locate an operand. Methods include immediate (operand is part of instruction), direct (address of operand in instruction), indirect (address of address), and others.

Stack Organization

• A stack is an ordered list where elements are added and removed from only one end (top of stack).
• It follows the Last-In, First-Out (LIFO) principle.
• Stack pointer register (SP) stores the address of the top of the stack.

Memory Organization

• Memory organization can be simultaneous (all levels directly connected to CPU) or hierarchical (levels accessed sequentially based on hit rate).

Cache Memory

• Cache memory is a specialized fast memory that stores frequently accessed data.
• Mapping algorithms determine where data from main memory is stored in the cache.
• Methods include direct mapping, associative mapping, and set-associative mapping; each has different characteristics (e.g., cost, complexity).

Cache Replacement Policies

• Replacement policies are used when the cache is full and a new block needs to be loaded.
• Common techniques include FIFO (First-In, First-Out), Optimal, LRU (Least Recently Used), MRU (Most Recently Used).

Types of Miss

• Compulsory miss occurs when the CPU reads a block from main memory for the first time.
• Capacity miss occurs when the cache cannot hold all blocks needed for a program.
• Conflict miss occurs in set-associative or direct-mapped caches when multiple blocks map to the same set or block frame.

Floating-Point Representation

• Floating-point numbers use a special representation storing a number as a mantissa and exponent.
• The exponent is typically stored in biased form.

IEEE 754 Standard

• The IEEE 754 standard provides the formalization for floating-point arithmetic.
• It defines conventions for accurate and consistent handling of these numbers in different hardware systems.

Instruction Format

• Instructions are classified into types based on the number of operands they use (e.g., 3-address, 2-address, 1-address, 0-address).

Micro-Operation

• A micro-operation is a small, atomic step in a sequence of operations. The operations are typically classified into groups such as register transfer, arithmetic, logic, or shift.

I/O Management

• I/O devices are independent devices that serve to communicate with the computer. The methods for data transfer include programmed I/O, interrupt-driven I/O, and direct memory access (DMA).

I/O Processor

• I/O processors are separate processors that handle input/output operations and communication with I/O devices, reducing the load on the main CPU.
• It is separate to the main CPU so that the data can be passed freely without affecting the main CPU.

Synchronous and Asynchronous Data Transfer

• Synchronous data transfer uses a clock signal to synchronize operations between devices, while asynchronous communication uses handshaking signals, which are slower but good for varying speeds.

Addressing Modes, Types and Examples

• Describes different methods for specifying the location of data operands. Includes Immediate, Direct, Indirect and others.

Types of Interrupts

• Hardware interrupts are initiated by external devices, while software interrupts are instructions embedded in a program. Interrupt processing includes recognition, status saving, masking, acknowledgment, service routine execution, and restoration.

Direct Memory Access (DMA)

• DMA is a way for I/O devices to directly transfer data to main memory without continuous CPU intervention or instructions.

Uniprocessing and Multiprocessing

• Uniprocessing involves a single CPU executing instructions sequentially, whereas multiprocessing allows multiple CPUs to handle multiple tasks concurrently.

Pipelining

• Pipelining achieves concurrent instruction execution. Operations are divided into stages or phases, and multiple instructions are processed concurrently in different stages at the same time.

Control Word Sequencing

• Control word sequencing in microprogramming involves fetching control words to specify control signals for executing instructions.

Memory Mapping, Hierarchy and Usage

• Memory is organized in a hierarchy with varying levels of speed, capacity, and cost to organize and store data required by a system. Includes registers, cache memory, main memory, secondary memory.

Replacement Algorithms, Locality of Reference and Cache Mapping

• Replacement algorithms determine how a cache chooses which blocks to replace when a new block must be brought into a full cache. Locality of Reference is very important. Cache mapping determines where data from main memory is stored in the cache. This includes the methods like Direct, Fully Associative, and Set Associative mapping.

Other Topics

• The document has additional sections on specifics of the memory, different cache mapping methods (Direct, Set-associative, Fully-associative), and details of different types of ROM (e.g., masked ROM, PROM, EPROM, EEPROM, flash memory).
• There are details on the specific characteristics of CISC and RISC types of architecture in terms of instructions and designs.