Chapter 3 discusses the technology used to input, process and output data. Hardware includes any machinery that participates in the input, output, processing, storage, or transmission activities of an information system. Most all hardware today consists of digital circuits. Hardware decisions are based in an understanding of how the hardware can be used to to support business goals.
After completing Chapter 3, you should be able to accomplish the objectives on the next 2 slides.
The objectives and characteristics of a computer system support those of the information system and organization of which it is a part. An understanding of the relationship of the computer system to the information system, and ultimately, to the business, is vital to selecting or assembling an effective computer system.
Information systems add value to organizations, yet their use is strongly influenced by the structure and culture of the organization.
Not only must current business and IS needs be considered when selecting computer hardware devices, but also future plans. Computer hardware should be scalable or adaptable to future needs.
A computer system is a set of integrated devices that input, output, process, and store data and information. Computer systems are currently built around at least one digital processing device. There are five main hardware components in a computer system: the central processing unit, or CPU; primary storage, or main memory; secondary storage; and input and output devices.
Program instructions are executed in the CPU. These instructions are not equivalent to a line of code in a program written by a programmer; each line of code may be made of numerous internal program instructions.
The machine cycle includes the activities involved in executing an instruction. The machine cycle consists of two parts: the instruction phase followed by the execution phase. During the instruction phase, the control unit retrieves the instruction to be executed from main memory. The instruction is then decoded so the CPU understands the work to be done. Necessary data is then retrieved from memory and stored in a register.
During the execution phase the arithmetic logic unit performs the instruction and the results are stored in registers.
Most central processors use pipelining to speed processing. During pipelining, as one instruction is being executed, another is being decoded, and yet a third is being retrieved. Modern chips often used more than one pipeline. For instance, the Pentium processor uses two pipelines and thus can execute two instructions during one machine cycle.
The system unit houses the processing components of the computer system. All other computer system devices are called peripherals, and are connected directly or indirectly into the system unit. Speedy and efficient processing is important. There are several ways to measure processing speed and many factors affect it.
The machine cycle occurs very quickly. The length of time of a machine cycle is measured in fractions of a second. These units include: micro seconds, one millionth of one second; nanoseconds, one billionth of one second; or picoseconds, one trillionth of one second.
Machine cycle time can also be measured by how many instructions are executed during a second. MIPS,or millions of instructions per second, is often used to measure the processing speed of computers. In the not too distant past, MIPS was a measure used only for large computers, such as mainframes or supercomputers; however, today, even the processing speed of laptop computers is measured in MIPs.
The CPU produces electronic pulses at a predetermined and constant rate. This is called the clock speed. Clock speed is generally measured in megahertz, that is, millions of cycles per second.
The control unit follows microcode to control the machine cycle. Microcode is predefined, primitive, internal instructions that the processor performance during the instruction cycle.
Recall that data exists as electrical voltages in a computer. Since electricity can exist in 2 states, on or off, by FLV Player Addon" style="background-color: transparent !important; border: none !important; display: inline-block !important; float: none !important; font-style: normal !important; font-variant: normal !important; font-weight: normal !important; font-size: 14px !important; line-height: normal !important; font-family: Verdana, Geneva, sans-serif !important; height: auto !important; margin: 0px !important; min-height: 0px !important; min-width: 0px !important; padding: 0px !important; vertical-align: baseline !important; width: auto !important; text-decoration: underline !important; background-position: initial initial !important; background-repeat: initial initial !important;">binary digits are used to represent data. Binary digits, or bits, can be “0” or “1”.
A computer word is the number of bits the CPU can process as a unit at one time. A bus, or bus lines, are the electrical connections between the CPU and other system components. The bus width is the number of bits that can move along the bus at once. For optimal speed, a computer’s bus width should equal its wordlength. If a bus is narrower than the wordlength, then data will be transferred more slowly than it is processed, thus slowing overall speed.
Thus, many factors affect processing speed, including clock speed, machine cycle time, wordlength and bus width. Thus, different processors cannot be compared directly. Several chip benchmarks have been developed to enable comparisons.
The speed of a processor is also limited by physical constraints. CPUs are made of digital circuits on silicon wafers, called chips. Electrical current flows through silicon to turn these circuits on and off. Since electricity travels at about the speed of light, moving these circuits even slightly closer together can greatly improve processing speed. New manufacturing techniques have resulted in continually smaller chips, with circuits more closely packed together. In the 1960’s, Gordon Moore, former Intel chairman of the board, stated that transistor (circuit) density of single chips will double every 18 months. Gordon’s Law remains true today, despite continual warnings that chip density will plateau due to the limitations of silicon.
Although Moore’s law continues to hold, the limits of silicon for processor construction will inevitably be reached and more transistors won’t be able to be placed on a silicon chip. Researchers are currently looking for an alternative to silicon. Superconductive metals that allow current to flow with minimal resistance is one alternative being explored, as is galium arsenide. Some firms are exploring the use of light instead of electrical current to represent bits.
Processing speed can also be improved by changing characteristics of the microcode instruction set. Originally, computers contained as many microcode instructions as possible. However, researchers found that the 80/20 rule applies to microcode: only 20% of the instructions are used 80% of the time. Since the 1980’s, RISC chips have included fewer microcode instructions and have been faster and less expensive than complex instruction set computers. Motorola’s PowerPC chip is an example of a RISC microprocessor.
An alternative approach to improving processing speed is to lengthen each instruction, instead of reducing their number. This is the approach taken in the design of very long instruction word chips.
Memory, also called primary storage, is located physically close to the CPU to decrease access time, that is, the time it takes the CPU to retrieve data from memory. Memory temporarily holds instructions and data before and after processing by the CPU. Although the overall trend has been increased memory access time, memory has not advanced as quickly as processors. Memory access time is often measured in milliseconds, or one thousandths of a second.
Like the CPU, memory is made of silicon chips containing circuits holding data represented by on or off electrical states, or bits. Eight bits together form a byte. Memory is usually measured in megabytes or gigabytes.
A kilobyte is roughly 1,000 bytes. Specialized memories, such as cache memories, are typically measured in kilobytes. Often both primary memory and secondary storage capacities today contain megabytes, or millions of bytes,of space.
Increasingly desktop computers come with gigabytes or billions of bytes of storage capacity on their hard disks or secondary storage. Although today terabytes of storage are found only in large computers such as mainframe computers, it shouldn’t be long before we are accessing terabytes of storage on our desktop computers.
In fact we are starting to hear of even larger quantities of storage. For example, there are operating systems that can access files that contain over one exabyte – one quintillion bytes - of data, and there are databases that hold over an exabyte of data. An exabyte equals about one quintillion bytes.
In addition to ROM, or read-only memory, programmable ROM and erasable programmable ROM exist.
IN ROM chips, the contents, or combination of electrical circuit states, are set by the manufacturer and cannot be changed. States are permanently manufactured into the chip.
In PROM, the settings must be programmed into the chip. After they are programmed, PROM behaves like ROM – the circuit states can’t be changed. PROM is used when instructions will be permanent, but they aren’t produced in large enough quantities to make custom chip production (as in ROM) cost effective. PROM chips are, for example, used to store video game instructions.
Instructions are also programmed into erasable programmable read-only memory. However, the contents of the chip can be erased and the chip can be reprogrammed. EPROM chips are used where data and instructions don’t change often, but nonvolatility and quickness are needed. The controller for a robot arm on an assembly line is an example of EPROM use.
Multiprocessing is the simultaneous execution of more than one instruction. Multiprocessing takes different forms. A coprocessor is an additional processor in a computer system that handles specific functions for the CPU, allowing the CPU to perform other processing activities. Coprocessors may be internal or external to the CPU. Common coprocessors include math coprocessors that handle complex mathematical calculations and graphics coprocessors that speed the manipulation of images.
การแปล กรุณารอสักครู่..