Multimedia Hardware

Ms P Kadebu
 Hardware

 Multimedia computers
 Video and graphics, audio, and networks
 CD and DVD
 USB and FireWire
 Processors
 Architecture

 Most computers support multimedia

 They are usually based on bus architecture
 Types of Hardware
 If you are asked to develop a multimedia project, you
require a fast computer with lots of speed and storage.
 There are many more things that you need to know
like which component makes a computer fast, what is
the device for storage, etc.
 The components are thus divided into five categories
1. System devices,
2. Memory and storage devices,
3. Input devices,
4. Output devices, and
5. Communication devices.
 System devices
 These include microprocessor, motherboard and memory.
 Microprocessor is basically the heart of the computer and it is a
computer processor on a small microchip.
 The microprocessor gets the first instruction from the
microprocessor
 Basic Input/Output System (BIOS), which is a part of its memory.
 BIOS actually loads the operating system into random access
memory (RAM).
 A Motherboard is a device in the computer that contains the
computer's basic circuitry and other
 components.
 Motherboard contains computer components like microprocessor,
memory, basic input/output system (BIOS), expansion slots and
interconnecting circuitry
 Memory and Storage devices
 RAM (random access memory), also called primary memory, locates
the operating system, application programs, and data in current use so
that the computer's processor reaches them quickly.
 RAM is called "random access" because any storage location can be
accessed randomly or directly. RAM is much faster than the hard disk;
the floppy disk and the CD-ROM. RAM can be taken as short-term
memory and the hard disk as the long-term memory of a computer.
However, RAM might get slow when used to its limit. That is why,
more memory is needed to work on multimedia. Today's personal
computers come with GBs of RAM.
 Hard disk stores and provides access to large amounts of data on an
electro magnetically charged surface.
 Compact Disk (CD) is a small medium that can store data pertaining
to audio, video, text, and other information in digital form.
 Compact Disk
 CD-ROM is well suited for multimedia distribution. It has
many standards whose specifications are defined in the
Rainbow Books:
 CD-DA (red book): Digital Audio, 150 kbps, 44.1
ksamples/s, 682 MB
 CD-ROM (yellow book):
› 1) text & data,
› 2) audio & video data
 CD-I (green book): text, graphics, audio, and video repeat
(multimedia and games). Compact Disk- Interactive
 CD…..cont’d
 CD-XA (extended yellow book): improves audio
and video
 PhotoCD: max 100 compressed pictures
 Video CD (white book): MPEG1 video (74 min)
 CD-R (orange book): write once CD. In the CD-R,
once the data recording is completed, it becomes a
CD-ROM and nothing can be deleted.
 CD-RW: re-writable CD. One can write, erase and
again write data into the same disk
 DVD
 Digital Versatile Disc
 DVD have bigger storage capacity than CD
1) smaller pit distance (0.4 microns vs. 0.834 microns)
2) smaller track distance (0.74 microns vs. 1.6 microns)
3) smaller laser wave length (635-650 nanometers vs.
780 nanometers)
4) two sides
5) two layers
 DVD capacity
 DVD capacity:
1) one side and one layer: 4.7 GBytes
2) two sides and one layer: 9.4 GBytes
3) two sides and two layers : 17 GBytes
 DVD drives can be compatible with CD records
 Video uses MPEG-2 compression and audio
Dolby AC-3
 DVD encryption

 Part of content is encrypted (saves processor time)

 DVD player uses hardware for decoding
 The encryption key is on the DVD disk – only
license holders know how to locate it
 TV signal is scrambled, so it can not be copied
 DVD-ROM drive gives encryption key via secure
bus (Content Scrambling System - CSS)
 Water stamps can be used for limited copying
 Next Generation DVD
 Blue laser (450 nm) instead of red (650 nm)
 Blu-Ray
› Sony, Hitachi, LG, Matsushita, Philips, Pioneer,
Samsung, Sharp, and Thomson
› 23, 25 tai 27 GB
 DVD-HD
› Toshiba, Nec, Warner, Paramount, and Universal
› 15-20 GB
 USB

 Universal Serial Bus (USB) can be used for

connecting different kinds of devices
 USB requires the presence of a bus master, typically
a PC, which connects point to point with the USB
slave
 USB advantages:
1) hardware settings (e.g., interrupts, DMA, addresses, etc.)
do not have to be defined
2) no need to “open” computer
3) no separate power wires
 USB devices

 keyboard, mouse, printer, modem, scanner, CD-

ROM, etc.
 Telephone, ISDN, etc.
 Joystick, game controllers, data gloves, head
displays, etc.
 Video cameras, digital cameras, etc.
 TV and radio receivers
 USB features
 USB speed is 1.5 Mbps (Low Bandwidth) and 12
Mbps (High Bandwidth)
 USB has four wires
 Bits are defined as voltage differences
 Maximum distance is 5 meters
 Devices are connected into star topology (HUB)
 Networks can be hierarchical
 USB supports

 Isochronous ( processes that require timing

coordination to be successful) video and audio
transfer
 Real-time interrupts
 Large data transfers with best effort method
 Control information transfer
 USB 2.0
 Transfer capacity max 480 Mbps (40 times)
 Compatible with USB 1.1
 Speed is dropped automatically if required
 Allows the use of new hardware:
 video conferencing (better image quality) printers
& scanners (speed & resolution)
 CDROM & DVD (more data)
 FireWire

 As USB, but larger transfer rate: 98.304,

196.608, and 392.216 Mbps
 Over 1 Gbps in development
 64 000 devices and 1 000 bridges
 One cable max distance 4.5 m
 Two devices max distance 16 cables (i.e., 72 m)
 FireWire configuration

 Device groups are linked via bridges

 FireWire cable
 Cable has two data wires and two power
 FireWire advantages

 Scales well
 Flexible topology
 Devices can be attached to a “hot” bus
 Easy to use (no settings)
 Designed for consumer devices (cheap)
 Peer to peer. No need for PC as host, devices can
connect to each other independently
 FireWire situation

 Limited amount of devices

 FireWire cards available for computers
 Some workstations and PC have FireWire as
standard feature
 Some video cameras, etc.
 A standard feature in future digital television set-
top-boxes
 Input devices
 Keyboard is the primary text input device for your
computer.
 Mouse is a small device that you move across a pad in order
to point to a place on a display screen and thus execute a
command by clicking it.
 Microphone is another input device that can interpret
dictation and also enable us to input sound like the keyboard
is used for text.
 Digital camera records and stores photographic images in
digital form that can be fed to a computer as the impressions
are recorded or stored in the camera for later loading into a
computer. The digital cameras are available for still as well
as motion pictures
 Output devices
 Printer is a device, which on receiving the signal from computer
transfers the information to paper. Earlier the Dotmatrix printer was
a popular low-cost personal computer printer; now inkjet printers
have taken its place. Dot-matrix printer strikes the paper a line at a
time while inkjet sprays ink and laser printer uses a laser beam to
attract ink (also called toner).
 Monitor is a device for display. It is just like a television set and is
measured diagonally from two opposing corners of the picture tube.
The standard monitor size is 14 inches. Very large monitors can
measure 21 inches diagonal or greater.
 Amplifier is an electronic device that increases the power of a
signal and is used in audio equipment.
 Speakers with built-in amplifiers have become an integral part of
the computers today and are important for any multimedia project to
output audio.
 Processors
 Multimedia processing requires performance
 Lot of data and complex algorithms
 Problem can be solved with:
1) hardware accelerators (MPEG decompression, etc.)
2) special processors (signal processors, etc.)
3) Extensions of standard processors eg MMX (single
instruction, multiple data (SIMD) instruction set), etc.
4) Parallel architectures
 Accelerators
 In accelerators, the algorithm is directly
implemented in hardware
– Data processing is fast Algorithms can be either
permanent or reconfigurable
– In latter case, re-programmable logic circuits are often
used
 Accelerator can be independent unit or it can be
attached to a host processor
– Accelerator can be even integrated to the same chip
with the host processor
 Special processors (I/II)

 Digital signal processors can be used

especially for image and audio signal
processing
 The instructions set contain common operations
used in signal processing
 Manufacturers often have also special algorithm
libraries for signal processing
 Many multimedia terminals have a signal
processor in addition to host processor
 Special Processors (II/II)

 Another common group of special processors is

graphics processors
 They are optimized for different drawing,
lighting, and rendering algorithms
 Graphics processors are used together with host
processors
Nvidia GeForce 6800
 MMX

 MMX is Intel Pentium and Pentium Pro

multimedia extension
 Based on Single Instruction Multiple Data
(SIMD) architecture
 One instruction control several data units at the
same time E.g., 64 bit register has eight 8 bit data
units or four 16 bit data units
Example: parallel multiply
 MMX in practice
 Same operation is repeated for several data units
 Video compression, image processing, computer
graphics, modems, and digital signal processing
 For example, MPEG-1 video decompression is
1.5 times faster than without MMX instruction
set
 The algorithms have to be recoded
 Parallel Architectures
 Processing power can also be increased with
parallel architectures:
› Superscalar architectures
› Vector processing
› Single Instruction Multiple Data (SIMD)
› Very Large Instruction Word (VLIW)
› Transport Triggered Architectures (TTA)
 Superscalar Architectures
 Implements instruction-level parallelism within a
single processor
 Processor has one instruction unit, but multiple
functional units Instruction unit distributes the
operations to different functional units for execution
 The order of the operations can he rescheduled
 Either in compiler or hardware
– Source code does not need rewriting
 Most powerful modern processors are superscalar
architectures
 Vector Processing
 Traditional super computers are based on vector
processing
 The data is stored as vector or matrix data
structures
 More than one data unit is processed
simultaneously in different functional units
 Functional units can be pipelined
 Vector processing requires either hand coding or
optimizing compiler
 SIMD

 In SIMD architecture, same instruction is

executed at the same time for many data units
 Architecture consists of one instruction unit and
several identical execution units
 Usually, data is first distributed to the execution
unit and collected after computing
 Fits well with image processing and computer
graphics
 VLIW

 In Very Long Instruction Word architectures , one

instruction contains operations for several
functional units
 Compiler takes care of building the instruction
words E.g., loop structures can be unrolled as
parallel operations
 Multimedia and signal processing algorithms
have often many tight loops
 TTA
 Transport Trigger Architectures do not have
instruction unit at all (smaller energy consumption
and chip area)
 Instead there are several functional units
 Compiler defines how data is transferred from one
functional unit to another
 Transfer is done as soon as computing is ready
 Fits well with signal processing
TTA Example