Computer Architecture and Organization Review

Flashcards covering memory types, access methods, CPU cycles, I/O transfer modes, system performance, and computer history from the lecture notes.

Primary Memory

A type of memory that is very fast and volatile (data lost when power is off), directly accessible by the CPU, including examples like RAM, Cache, and Registers.

Secondary Memory

A type of memory that is slower and non-volatile (data retained when power is off), having large capacity (GBs-TBs) and accessed indirectly via primary memory.

Sequential Access

An access method where data is accessed in a fixed, linear order, one after another, as seen with magnetic tape.

Direct Access

An access method where data is stored in blocks, allowing the system to jump directly to a block without reading everything before it.

Random Access

An access method where any memory location can be accessed directly and instantly, regardless of its position, used in RAM and SSDs.

Fetch

The CPU process of reading the next step in the recipe or instruction from the memory.

Decode

The CPU process of understanding what a specific instruction step means.

Execute

The CPU process of performing an action such as chop, mix, or boil (mathematical or logical operations).

The Power Wall

A problem encountered around 2005 where increasing clock frequency led to unsustainable heat and power consumption.

Parallel Processing

The act of dividing a task into sub-tasks that execute simultaneously using multiple simpler cores on a single chip.

Amdahl's Law

A principle that dictates the potential speedup gained by using multiple processors.

Instruction Set Architecture (ISA)

The critical interface between user/software and hardware, defining attributes visible to the programmer.

Programmed I/O

An I/O method where the CPU is responsible for checking the status of the I/O device periodically through a mechanism called polling.

Burst Mode

A DMA mode where the controller takes control of the bus and transfers an entire block at once, blocking the CPU during the transfer.

Cycle Stealing

A DMA mode where the controller forces the CPU to pause for just one bus cycle to transfer one word, interleaving with CPU execution.

Transparent Mode

A DMA mode where data is transferred only when the CPU is not using the bus.

Embedded Systems

Dedicated functional computers within a larger system, such as microcontrollers in cars or appliances.

System on a Chip (SoC)

An integrated circuit that combines all components of a computer, including the CPU, memory, and I/O, into a single chip, ubiquitous in smartphones.

Response Time (Latency)

A performance metric measuring how long it takes to do a task, typically in $\text{seconds/program}$.

Throughput (Bandwidth)

A performance metric measuring how many tasks are completed per unit of time, typically in $\text{programs/second}$.

CPU Execution Time

A performance formula calculated as $\text{Instruction Count} \times \text{CPI} \times \text{Clock Cycle Time}$.

I/O Processor (IOP)

A specialized processor with its own memory and instruction set dedicated to I/O tasks, capable of executing its own Channel Programs.

Von Neumann Model

A computer model proposed in 1945 featuring a stored-program concept where instructions and data are stored in the same memory.

Von Neumann Bottleneck

The limitation where sequential instruction processing is restricted by the speed of the memory bus.

Interrupt-Driven I/O

An I/O method where the CPU issues a command and continues other tasks until the I/O module sends an Interrupt Request (IRQ).

First Generation Computers

Computers developed between 1946-55 using vacuum tubes, such as the ENIAC.

Second Generation Computers

Computers developed between 1955-65 using transistors, such as the IBM 7000 series.

Third Generation Computers

Computers developed between 1965-80 using Integrated Circuits (ICs), such as the IBM System/360.

Fourth Generation Computers

Computers developed from 1980 to the present using VLSI Microprocessors housing thousands to billions of transistors.

Computer Architecture

The 'What' of a system, referring to attributes visible to a programmer, such as instruction sets and data type bits.

Computer Organization

The 'How' of a system, referring to operational units and interconnections like control signals and bus widths.

Direct Memory Access (DMA)

A technique where a specialized hardware module transfers data directly between I/O and memory without CPU intervention.

Moore's Law

The 1965 observation that the number of transistors on a chip doubles approximately every two years.