Computer Organization &

Architecture
 Computer Organization

• Computer organization is concerned with the way the hardware

components operate and the way they are connected together to
form the computer system. A knowledge of computer organization
can help one to understand the internal operations that are carried
out by a computer while a program is being executed.

• The various components of a computer are assumed to be in place

and the task is to investigate the organizational structure to verify
that the computer parts operate as intended.
 Computer Organization

• We can therefore say that computer organization

deals with the internal view of the computer and
the roles that the internal components play
during execution of a program.
 Computer Architecture

• The architecture (also called instruction set architecture) of a

computer is concerned with the structure and behavior of the
computer as viewed by user such as an assembly or machine
language programmer.

• An assembly or machine language programmer needs to be aware of

the specific instructions supported by the processor (called the
processor’s instruction set), the instruction formats, the specific
registers and their roles, the techniques for accessing the data stored
in the memory, the way to perform input/output operations, etc.
 Computer Architecture

• Thus, computer architecture is the external view

of a computer that is essential to be properly
understood by anyone who is likely to program a
computer using machine or assembly languages.
 Modules to be covered
•Introduction and Performance Evaluation
•Data representation and basic operations
•Instruction Set Architecture
•Processor Design
•Memory hierarchy
•Input/Output Organization
•Parallel Processing
 Introduction and
Performance Evaluation
 Introduction

• Digital computers use the binary number system, which has two
digits: 0 and 1.

• A binary digit is called a bit.

• Information is represented in digital computers in groups of bits.

• By using various coding techniques, groups of bits can be made to

represent not only binary numbers but also other discrete symbols,
such as decimal digits or letters of the alphabet.
 Introduction
• By judicious use of binary arrangements and by
using various coding techniques, the groups of
the bits are used to develop complete sets of
instructions for performing various types of
computations.
 Why we need to Study Computer
Organization and Architecture?
• To get an insight of how a computer executes programs internally.

• To write more effective programs.

• To understand why a program exhibits poor performance, even

though it might be implementing a computationally efficient
algorithm.
 Why we need to Study Computer
Organization and Architecture?

For system programmers and programmers of embedded

computing systems, a good knowledge of computer
architecture and organization is essential because they often
need to program bare hardware without the support of any
operating system.
Organization implements the architecture
This figure shows the relation between
co m p u te r o rga n i zat i o n , co m p u te r
architecture, system programs , and
application programs.

It also shows that system program directly

interacts with the computer hardware.

The system programs are written for

specific computer architectures.

The application programs on the other

hand invoke the services offered by the
system programs.
 Computer design

• Computer organization is an implementation of the architecture of a

computer .

• Starting from the specification of the architecture of a computer, and

based on several constraints that may have to be satisfied, the
organization is arrived through computer design.

• Computer design is concerned with the hardware design of the

computer.
 Computer design

• Once the architecture and other specifications of the computer are

formulated it is the task of the designer to develop the hardware for
the system.

• Computer design is concerned with the determination of what

hardware to be used and how they are to be connected. This aspect
of computer hardware is sometimes referred to as computer
implementation.
 Von Neumann Computers

• In the very early days of computing, one instruction at a time of a

program was executed, with an operator setting up each instruction
and also initiating the execution of each instruction .
• The operator would typically set up an instruction by flipping some
switches on switch board, and then would set up the required data
and control paths to the functional units by using patch cords.
• Certainly, executing a program was extremely cumbersome and
required a lot of effort by the operators.
• At this juncture, in a revolutionary step, the concept of stored
program computers was proposed by John Von Neumann in 1945.
 Von Neumann Computers

• Von Neumann proposed that the instructions can be encoded and

stored in the memory just like data.

• During the execution of a program, the stored instructions can be

fetched from memory , and the fetched instruction can be decoded to
set up the necessary data paths and generate the control signals.
 Organization of Von Neumann computer

The organization of a Von Neumann computer is shown schematically in the below

figure.
The figure shows that
a control unit is
incorporated as a part
of the processor.
 Von Neumann Computers

• The control unit is an electronic circuitry that decodes each fetched

instruction, and generates the necessary control signals that are
transmitted to the functional units to carry out the required
operations and also the control signals are responsible to
automatically set up the required data paths.

• In essence, the stored program concept along with the control unit,
eliminated the need for an operator to enter each instruction through
switch settings, and also eliminated the need to set up data paths by
inserting patch cords.
 Von Neumann Computers

• Von Neumann’s novel computer organization made the program

execution much faster.
 Von Neumann Computers

• Execution of programs became faster as manual entry of the

instructions as well as setting up of the data and control paths to the
functional units could be totally avoided. This gave rise to the concept
of an instruction cycle.

• In an instruction cycle, an instruction is fetched from memory,

analyzed, the data fetched, processed and finally the results are
produced. In other words, the execution of an instruction is
completed in an instruction cycle.
 Shortcomings of Von Neumann computers.
• Towards the end of 1980s, the quest for faster computations made
designers to notice a few shortcomings of the Von Neumann
computing.

• A shortcoming is the existence of a single connection between the

processor and the memory.
• This suggests that at a time only one memory access can occur.
• Hence, at a time either an instruction can be fetched or a data can be
accessed. This appeared as the problem in parallel execution of
instructions.
 Shortcomings of Von Neumann computers.

• The term Von Neumann bottleneck is often used to indicate that both
fetching an instruction and accessing (reading and writing) data over
the same bus from memory by the processor at the same time in a
Von Neumann computer is not possible , this is a bottleneck in
achieving high-performance computations.
 Basic Organization of a computer

• The important parts of a digital computer are the processor, main

memory, hard disk (secondary memory), key board, monitor, and
peripheral devices.

• In simplest organization of a computer, all these different parts of a

computer can be connected trough a single bus called backplane bus.
 Basic Organization of a computer

A single bus interconnecting all the components of a computer is

usually called a backplane bus. This is so because the single bus can be
considered to be a backbone communication medium, to which various
components are attached.

Although a backplane was used in the early computers, it was later

replaced by multiple specialized buses in modern computers for
achieving higher performance
Basic organization of a digital computer
 Basic organization of a digital computer
Processor
The processor or the central processing unit (CPU) is responsible for
fetching an instruction stored in the memory and executing it.

It contains an arithmetic an logic unit for manipulating data , a number

of registers for storing data, and a control unit.

The control unit generates the necessary control signals for fetching and
executing the instructions, and storing the results.
 Basic organization of a digital computer
Main Memory
The main memory is also called random access memory (RAM)
because the CPU can access any location in memory at random
and either retrieve the binary information stored at that
location, or can store some binary information at the location.

The main memory at present is usually made from dynamic

RAM (DRAM) modules.
 Basic organization of a digital computer
Other devices are
Monitor
Keyboard
Peripheral devices – Several peripheral devices can be
attached to the backplane bus. These peripheral devices can be
either output devices, such a printers and loud speakers, or can
be input devices, such as scanners and cameras. The CPU can
provide output to these output devices or collect input from
the input devices over the backplane bus.
Basic organization of a digital computer

Backplane bus
The backplane bus is a group of wires. These wires are partitioned into
control and address wires. The control wires carry control signals to
different units. The address wires carry the address of the specific data
in memory and the data wires are used to carry the data .
 Historical Perspective

First generation computers (1941 - 1956)

Second generation computers (1956 - 1963)

Third generation computers (1964 - 1971)

Fourth generation computers (1971 - Present)

Fifth generation computers (Present and Beyond)

 Performance Benchmarking

• Before we study the organization and architecture issues, we need to

understand how one can determine whether one processor (or
computer) works faster than another .

The reasons can be:

• Program execution is faster
But this knowledge may be inadequate in many situations.
• Next, Comparison of clock rates
 Performance Benchmarking

Execution time = CPI X clock cycle time X number of instructions.

CPI – Average clock cycles per instruction.

Clock cycle time is the inverse of clock rate.

This expression indicates that the performance of a computer depends

on not only its clock rate but also its CPI.
 Performance Benchmarking

Therefore, comparing the clock rates of two processors to determine

their relative performance would be a meaningless exercise.

About two decades ago, MIPS (millions of instructions processed per

second) and MFLOPS (millions of floating-point operations processed
per second) became popular measures for performance evaluations
and comparisons. However, there were shortcomings of this.
 Performance Benchmarking

The first problem is that the MIPS rating is instruction set dependent.

Second, the MIPS rating of the same processor would vary depending
which program is selected for running on the computer to determine
the MIPS rating. This is because MIPS rating would depend on whether
the program consists of very simple instruction or more complex
instructions.
 Performance Benchmarking
• Later synthetic benchmarks were proposed as a means to help
more accurately determine the performance of computers.

Main concept of synthetic benchmarks :

From a study of a large number of programs, the different types of instructions

(arithmetic, branch, memory access, etc.) that are used and their average
frequencies of occurrence are determined.

Based on this , a benchmark is a program that is synthesized to comprised a

sequence of instructions that has the same distribution of various types of
instructions as a typical program.
 Performance Benchmarking

That is, a synthetic benchmark program is just a collection of required

types of instructions in required frequencies of occurrence , but need not
compute any meaningful results.

Examples of synthetic benchmark programs are Dhrystone, Linpack,

Khornerstone, etc.
 Performance Benchmarking
The synthetic benchmark programs did get used extensively for some
time to determine computer performance.

However, the synthetic benchmark programs had an important

shortcoming that many times rendered the performance results
inaccurate.

All synthetic benchmark programs were rather short programs and

compiler writes could produce optimized code that artificially made
the computer work well for those programs only.
 Performance Benchmarking

• Later, benchmark suite (collection of benchmark programs) was

proposed as a solution.

• A popular benchmark suite is available from a nonprofit organization

called Standard Performance Evaluation Corporation (SPEC).

• The SPEC benchmark programs can be downloaded free of charge

from www.specbench.org
 Performance Benchmarking

• The available benchmark suite include : SPEC 2000 for computer

system performance, SPEC JVM98 for the performance of the client-
side Java virtual machine, SPEC JBB2000 for server-side Java
application, SPEC WEB2005 for evaluating web severs, etc.

• So, to determine which computer would work best as a web server

among the ones that meet the budget, would require comparing the
SPEC WEB2005 benchmark performance of the computers.
 Moore’s law
The one constant for computer designers is rapid change, which is driven largely by
Moore’s Law. It states that integrated circuit resources double every 18–24 months.

Moore’s Law resulted from a 1965 prediction of such growth in IC capacity made
by Gordon Moore, one of the founders of Intel.

As computer designs can take years, the resources available per chip can easily double
or quadruple between the start and finish of the project.

Like a skeet shooter, computer architects must anticipate where the technology will be
when the design finishes rather than design for where it starts.
Moore’s law
 Abstractions

Both computer architects and programmers had to invent techniques

to make themselves more productive, for otherwise design time would
lengthen as dramatically as resources grew by Moore’s Law.

A major productivity technique for hardware and software is to use

abstractions to represent the design at different levels of
representation; lower-level details are hidden to offer a simpler model
at higher levels. We’ll use the abstract painting icon to represent this
second great idea.
 Abstraction

Abstract painting
icon
 Clocking methodology
The approach used to determine when data is valid and stable relative
to the clock is clocking methodology.

Edge-triggered clocking A clocking scheme in which all state changes

occur on a clock edge.
 Clocking methodology

• A clocking methodology defines when signals can be read and when

they can be written.
• It is important to specify the timing of reads and writes, because if a
signal is written at the same time it is read, the value of the read
could correspond to the old value, the newly written value, or even
some mix of the two!
• Computer designs cannot tolerate such unpredictability. A clocking
methodology is designed to make hardware predictable.
 Clocking methodology

• For simplicity, we will assume an edge-triggered clocking

methodology. An edge-triggered clocking methodology means that
any values stored in a sequential logic element are updated only on a
clock edge, which is a quick transition from low to high or vice versa
(see Figure).
• Because only state elements can store a data value, any collection of
combinational logic must have its inputs come from a set of state
elements and its outputs written into a set of state elements. The
inputs are values that were written in a previous clock cycle, while the
outputs are values that can be used in a following clock cycle.
Clocking methodology
Combinational logic, state elements, and
the clock are closely related.

In a synchronous digital system, the clock

determines when elements with state will
write values into internal storage. Any
inputs to a state element must reach a
stable value (that is, have reached a value
from which they will not change until after
the clock edge) before the active clock
edge causes the state to be updated.
 Clocking methodology

All state elements in this chapter, including memory, are assumed to be

positive edge-triggered; that is, they change on the rising clock edge.
 Clocking methodology
Figure shows the two state elements surrounding a block of
combinational logic, which operates in a single clock cycle: all signals
must propagate from state element 1, through the combinational logic,
and to state element 2 in the time of one clock cycle. The time
necessary for the signals to reach state element 2 defines the length of
the clock cycle.
 Clocking methodology

For simplicity, we do not show a write control signal when a state

element is written on every active clock edge. In contrast, if a state
element is not updated on every clock, then an explicit write
control signal is required. Both the clock signal and the write
control signal are inputs, and the state element is changed only
when the write control signal is asserted and a clock edge occurs.
 Clocking methodology

We will use the word asserted to indicate a signal that is

logically high and assert to specify that a signal should be
driven logically high, and deassert or deasserted to
represent logically low. We use the terms assert and
deassert because when we implement hardware, at times
1 represents logically high and at times it can represent
logically low.
 Amdahl’s law
A rule stating that the performance enhancement possible with a
given improvement is limited by the amount that the improved
feature is used. It is a quantitative version of the law of diminishing
returns.
The execution time of the program after making the improvement is
given by the following simple equation known as Amdahl’s Law:

Execution time after improvement

 Amdahl’s law

• We can use Amdahl’s Law to estimate performance improvements

when we know the time consumed for some function and its
potential speedup. Amdahl’s Law, together with the CPU performance
equation, is a handy tool for evaluating potential enhancements.
Stored Program Architecture
Notion of IPC
Thank you