1. Course : AIT 204
Course Title : Computer Organization and Architecture (COA)
Course Credits : 3 + 0
Course Teacher : Dr. Y R Ghodasara & Prof. K.C. Kamani
College of Agricultural Information Technology
Anand Agricultural University
Anand
Unit IV
2. Intro to the I/O system
• I/O Devices are attached with the computer to communicate with the external world.
• Input devices are used to take input from the user.
• Output devices are used to display output to the user.
• Storage devices are used to store files.
• I/O Devices are connected with the south bridge of the mother board.
• Different I/O devices are using different I/O Bus for Connectivity.
• I/O bus is a standard mechanism to connect I/O device to transfer data.
• Separate Controllers are there to control I/O Bus.
3. Criteria North Bridge Bus South Bridge Bus
Devices CPU, RAM, AGP,
Ethernet
All other devices
Bus Options FSB
PCI Express
AGP
PCI
PCI Express
USB
ATA
SATA
SCSI
FireWire(WiFi)
Clock Speed 70-1000 MHz 10-40 MHz
Data Transfer Speed 3 GB/Second 20-500 MB/Second
Comparison of North Bridge and South Bridge Buses
4. Bus Type Devices
Serial, KBD, PS2 Keyboard, Mouse, Floppy
ROM,CMOS BIOS
ATA, SATA Hard disk, CD-ROM/RW, DVD
PCI/ PCI Express Network Card, Sound Card, Graphics and Video
Card and other Adapters
USB Mouse, Scanner, Printer, External Hard Disk,
Modam and many more
FireWire(Wi-Fi) Scanner, Camera, External Disk
SCSI Hard Disks, CD Drive, DVD Drive, Scanners
LPT, COM Keyboard, Mouse, Printers
Different Bus Interface to connect Devices
5. I/O devices Data Transfer Methods
Three methods are used to transfer data between memory and I/O devices.
• Programmed I/O transfer
• Interrupted I/O transfer
• DMA (Direct Memory Access ) transfer
6. Programmed I/O Transfer
In the programmed I/O method, the I/O devices does not have direct access to memory.
A transfer from an I/O device to memory requires the execution of several instructions by the CPU.
CPU transfers data from I/O devices to its register and then write those data from register to
memory.
I/O
Controller
CPU RAM
Issue Read
Command to IO
Controller
Read Status of IO
Controller
status
Read Data from IO
Controller
Write Data to Memory
done
Not Ready
Ready
No
Yes
Execute Next Instruction
Check status
7. Disadvantages
• Continuous CPU Monitoring is required to check whether data is available from IO device or not.
This wastes lot of precious CPU cycles.
• Transfer from IO controller to CPU and CPU to Memory requires CPU execution time.
• Not suitable for large amount of data transfer.
8. Interrupted I/O Transfer
In the interrupt I/O method, the I/O device controller will generate interrupt when device is ready to
transfer data. On receiving this interrupt, the CPU will execute IO routine stored in the memory to
transfer data and then resumes its former processing.
I/O
Controller
CPU RAM
Interrupt Received
Interrupt Raised
CPU doing something
else
IO Controller
Interrupt for
Data trasnfer
Interrupted
Read Data from IO
Controller
Write Data to Memory
done
No
Yes
9. Advantages
• Continuous CPU Monitoring is not required. IO controller request(interrupt) for CPU when ever
required. CPU can execute other programs in free time.
Disadvantages
• Transfer from IO controller to CPU and CPU to Memory requires CPU execution time.
• Not suitable for large amount of data transfer.
10. DMA (Direct Memory Access)Transfer
In the DMA transfer, the I/O device is instructed by the CPU to transfer data.
CPU also provide location information and number of bytes to be transferred.
IO Controller will transfer data directly in the Memory using data bus.
CPU is free during this transfer and can do execution of other program.
IO Controller will inform CPU after transfer is complete.
I/O Controller
CPU RAM
Transfer Data From
location x1 ( 100
bytes ) to
x2(Memory) Transfer Complete
CPU Inform IO
Controller to transfer
data from x1(size) to
x2
CPU
Starts
other work
IO Controller Reads
Data
Write Data to Memory
done
No
Yes
Inform CPU
that transfer
is over.
11. Advantages
• Continuous CPU Monitoring is not required. IO controller request(interrupt) for CPU when ever
required. CPU can execute other programs in free time.
• Transfer from IO controller to Memory does not require CPU execution time. CPU remains free.
• It is suitable for large amount of data transfer.
13. Industry Demands More Processing Power For
• Drug Design
• To simulate engineering model
• To general complex graphics in computer application
• To process large amount of historical data and predict future trend
• To predict financial and economical trends
Limitations of Single Processor Machine
• Clock speed can not be increase further.
• In Digital circuit maximum signal speed is as much as light speed.
• More speed generate more heat.
14. • Dual Core Chip uses one die in a single package and a die contains two core
units on the processor.
• Dual core has 4MB of shared L2 cache on the processor.
• On Intel dual-core CPUs the Front Side Bus is used for accessing the RAM
memory.
Dual Core Processor
Dual Core Processor
15. • Quad Core Chip uses two dies in a single package and each die is a dual core
unit, for a total of four cores on the processor.
• Each core still has 4MB of shared L2 cache for a total of 8MB L2 on the
processor.
• On Intel quad-core CPUs the Front Side Bus is used for accessing the RAM
memory and for the communication between each pair of cores.
Quad Core Processor
Quad Core Processor
16. Multiprocessor in Multi computers
Multiprocessor systems can be build by connecting hundred of computers with single
processor.
Computers co ordinate by various programming language facility like message passing.
This is a low cost solution to achieve 100 CPUs speed using ordinary computers.
Examples
Cluster Computing
Grid Computing
Cloud Computing
17. CLUSTER COMPUTING
A computer cluster is a group of tightly coupled computers
that work together closely so that in many respects it can be
viewed as though it were a single computer.
Clusters are commonly connected through fast local area
networks.
Clusters are usually deployed to improve speed and/or
reliability over that provided by a single computer, while typically
being much more cost effective than single computer the of
comparable speed or reliability .
19. 19
What is a Grid?
• Many definitions exist in the literature
• Early defs: Foster and Kesselman, 1998
“A computational grid is a hardware and software
infrastructure that provides dependable, consistent,
pervasive, and inexpensive access to high-end
computational facilities”
• Kleinrock 1969:
“We will probably see the spread of ‘computer utilities’,
which, like present electric and telephone utilities, will
service individual homes and offices across the
country.”
20. 20
3-point checklist (Foster 2002)
1. Coordinates resources not subject to
centralized control
2. Uses standard, open, general purpose
protocols and interfaces
3. Deliver nontrivial qualities of service
• e.g., response time, throughput, availability,
security
22. 22
Why do we need Grids?
• Many large-scale problems cannot be solved
by a single computer
• Globally distributed data and resources
• Roughly on a grid, a server log in to a bunch of
computers (the grid), send them data and a
program to run, and runs the program on
those computers, which sends the data back
to the server when its done.
23. Characteristics of Grid Computing
• Loosely coupled (Decentralization)
• Diversity and Dynamism
• Distributed Job Management & scheduling
24. 24
Comparison of Grid Computing vs. Cluster
Computing
• Grid computing is focused on the ability to support
computation across administrative domains sets it apart from
traditional computer clusters or traditional distributed
computing.
• Grids offer a way of using the information technology
resources optimally inside an organization. In short, it involves
virtualizing computing resources.
• Grid computing is often confused with cluster computing.
Functionally, one can classify grids into several types:
Computational Grids (including CPU scavenging grids), which
focuses primarily on computationally-intensive operations,
and Data grids, or the controlled sharing and management of
large amounts of distributed data.
25. Cloud computing
• Cloud computing is the use of computing
resources (hardware and software) that are
delivered as a service over a network (typically
the Internet). The name comes from the use
of a cloud-shaped symbol as an abstraction for
the complex infrastructure it contains in
system diagrams. Cloud computing entrusts
remote services with a user's data, software
and computation.
26.
27. Use
•Helps to use applications without installations.
•Access the personal files at any computer with
internet access.
•This technology allows much more efficient
computation by centralizing storage, memory,
processing and band width.
