Hardware for Processing data Processing hardware implements

Hardware for Processing data

Processing hardware implements the instructions encoded into software. A special hardware device called as instruction register or instruction counter contains the address of a location in the computer’s memory that holds the instruction a computer needs to start its operation.

Turning on the computer activates this instruction. Thereafter, the processor execute instructions sequentially in the order they exist in memory unless it encounters an instruction that resets the instruction counter to a different memory address.

Steps needed for a computer processor to perform its work Fetch instruction from memory Execute instruction Decode instruction

Measuring processor power Clock speed (тактовая частота) is the most common measure of processor power. An electronic circuit called a clock emits a regular electronic beat or pulse that synchronizes the operation of the processor.

Measuring processor power With each pulse, the processor performs one operation. Some operations, particularly the execution of complex instructions, may take several pulses. The term “hertz” is used to refer to the number of pulses c clock produces per second.

Prefixes use to modify measures of processor speed and quantity of storage Prefix Meaning Kilo One thousand Mega One million Giga One billion Tera One trillion Peta One quadrillion

Other measures Word length and Bus Width Speed of arithmetic Instruction speed instruction set Pipelining

Word Length and Bus Width The number of bits a computer can process at one time. It is usually at least 32 for PC and 64 for organisational computers. The Bus Width refers to the number of bits a computer can move at one time from one area of memory to another and is generally less than word length.

Speed of Arithmetic The flop (number of floating point operations per second) measures arithmetic power. Modern computers operate in the gigaflop to teraflop range.

Instruction speed MIPS (millions of instructions per second) measures instruction speed for a given processor type. Instruction speed usually varies directly with the clock speed.

Instruction set The number of different instructions a processor can decode and execute

Pipelining A processor’s ability to overlap the fetching, decoding, and executing of different instructions.

Parallel processing Uses two or more processors in one computer. These procesors either share a common bus and devices or operate more independently.

Specialized processors They respond to a limited set of commands to perform highly specialized tasks. Because their instruction have been “hard wired” into their chips and chips have been optimized for their tasks, they can perform these tasks more quickly than general-purpose processors.

Eg, Voice processors can translate sound-wave inputs into sound groups called phonemes and then into written words. They can increase the intelligibility and amplification of voice communications. They can also digitize and reproduce audio stored in computer files.

Hardware for storing data

Measuring storage Sequences of bits whose value can be zero or one, can represent all information. Bits can represent letters, numbers and other characters. Codes that use only seven bits can represent 128 different characters. Two coding systems that use eight bits have become industry standards.

ASCII code American Standard Code for Information Interchange Most PC use a code called ASCII to represent characters.

EBCDIC (Extended Binary Coded Decimal Interchange Code) IBM uses a code called EBCDIC for its largest computers. Some other manufacturers too.

Byte Because of the historical widespread use of these eight-bit coding schemes, we measure storage capacity in bytes, where one byte equals eight bits.

Companies Have been increasing or plan to increase their data storage by >33% a year. Some of them grow by 50% a year. Wal-Mart’s data warehouse held 7. 5 terabytes of the company inventory, forecast, customer, competitor, and market basket information in 2001.

Types of storage Primary or secondary Volatile or non-volatile

Primary or secondary Primary storage is electrical, resides on the bus, and is directly accessible to the processor. Secondary storage is storage that the processor cannot access directly. When the processor needs data, it commands the controller to obtain the data from the secondary storage device and place it on the bus. The processor then uses the data immediately or keeps it in primary storage.

Volatile storage or non-volatile storage Volatile storage requires electrical power to retain its data. Non-volatile storage retains its data even in the absence of electrical power.

Types and examples Primary RAM Cache Non-volatile ROM Secondary Volatile Disk Tape CD DVD Flash …

Access to data in secondary storage occurs at speeds several thousands to a million times slower than access to data in primary storage.

Advantages over primary storage 1. Much less expensive 2. Retains data without electrical power 3. Can be removed from its computer allowing the transfer of data between computers or the shipping of data as products

Cache memory Describes a small amount of primary storage that is faster than the rest of the primary storage in a computer.

ROM Read Only Memory retains its state in the absence of electrical power to hold the computer’s initial instructions. ROM devices do not change their state in response to an electronic signal; data must be burned into ROM memory using special equipment. Because of the expense of ROM and because data on ROM cannot be changed, computers contain only kilobytes of ROM compared to megabytes of RAM

RAM Random Access Memory loses whatever data it has if someone turns off the computer without any data or programs in its memory the computer could not do anything when turned on.

Secondary storage devices …

Distributed storage A storage area network (SAN) is a virtual storage device created by connecting different types of storage devices, over a high-speed network. A SAN centralizes storage so that any server connected to the SAN has access to its entire storage capacity.

RAID A technology called RAID (redundant arrays of inexpensive disks) uses a large number of relatively small hard disks to create what appears to be a single storage device. RAID reduces the time required to read or write data because the computer can simultaneously read or write to each of the disks in the RAID array. RAID storage also takes significantly less room that conventional large disks.

WAN – wide area network LAN – local area network

Hardware for Data Output Softcopy Hardcopy Robotic

Softcopy Output on an unmovable medium, such as a computer screen

Hardcopy Output on a medium, such as paper, that can be removed from the computer

Robotic Output into devices that physically move in response to signals from a computer

Density, resolution Density refers to the number of dots a device produces per inch horizontally and vertically 1, 000 dpi 640× 480. Resolution may be used to refer to density or dot pitch – the space between the adjacent dots.

Softcopy devices Display units Projectors Speakers

Display or screen Provides graphical visual output. Each dot, or pixel on the display corresponds to a location in the computer’s primary memory or in the memory on the video adapter – a circuit board inside the computer that supports the display.

projectors Allow computer images to be displayed to an entire room. Projectors can be compared on the basis of their brightness, contrast, resolution, and sharpness.

Speaker A sound controller, also called a sound card, produces an electrical signal that drives one or more speakers. This signal can produce music, special effects, such as the sound of a passing rain or the sound of a voice.

Hardcopy devices Produce output on media that can be removed from a computer and hence can be retained for a long time. The most common hardcopy medium for the computer is paper. And the most common output devices are printers and plotters. They are characterized by the density and resolution.

Printers Used by most individuals and businesses for the bulk of their output, produce text and graphics on paper without using a pen. Printers include laser, ink jet, matrix, and character impact.

Plotters Operate by moving a pen or pens over paper, much the way a person writes. Architectural and engineering firms often use plotters rather than printers to produce drawings. The plotters produce highresolution graphical output. They work on oversized paper and on long rolls of paper.

More devices Output onto Transparencies Microfilm Microfiche CD-ROM Slides

Robotic devices They physically move in response to signals from a computer.

Packaging computer hardware Manufacturers package the components of computers in different ways to meet different needs of markets. One way to characterize computer hardware is by its size: handheld, laptop, desktop.

Other types Computers to address the needs of an organisational enterprise are classified typically by their power and function

Mainframe computers Are the most powerful, typically running many application packages simultaneously and serving hundreds of users.

Midrange computers Are less powerful, typically serving a department

Servers No matter what their power, run a single application, such as a database or a Web connection. They may directly serve end users, but are just as likely to serve other computers running more general applications. Mainframes, midrange computers, and even personal computers can be operated as servers.

Blade server Is a computer constructed on a single circuit board that can be mounted on a specially constructed rack of similar servers rather than a computer box. Because of their low energy and space requirements companies use blade servers extensively in their data centers where a single rack may accommodate 200 or more such servers.

Key terms Adaptor Bit Byte Cache memory Data bus Density Flop Hard disk Hard copy

Mainframe MICR OCR Parallel processing Pixel Plotter Port Primary storage

RAID RAM ROM Resolution Robotic Secondary storage Sensor Server Softcopy Storage area network Volatile / non-volatile storage