Computer System Organization

Computer is a fast and accurate manipulating system. The term “system” is used in many ways. For computer users, a system is a group of parts that are integrated for the purposes of achieving some objective. The three following characteristics are key:

1. A group of parts. A system has more than one element.

2. Integrated parts. A logical relationship must exist between system parts.

3. A common purpose of achieving some objective. The system is designed to accomplish one or more goals. All system elements should be tied together and controlled so that the system goal is achieved.

Since a computer is a group of integrated parts that have common purpose of performing the operations called for in the program being executed, it qualifies as a system.

Computer system is the input, processor and output elements.

Computer systems use input devices – machines used for data entry purposes. Some machine devices allow direct human/machine communication, while others require data to be recorded on an input medium such as magnetizable material. Devices that read data recorded on a specially coated plastic flexible or floppy disks and magnetic tapes are often used. The keyboard of a workstation connected directly to a computer is an example of a direct input device. Devices connected directly to a computer are said to be online to it. The mouse, input pen, touch screen, and microphone are other online input devices.

Processor unit is the heart of any computer system, which consists of primary storage, arithmetic-logic, and control elements.

1. The primary storage section (also called main memory) section is used for four purposes. Three of these relate to the data being processed:

a. Data are fed into an input storage area where they are held until ready to be processed.

b. A working storage space that’s like a sheet of scratch paper is used to hold the data being processed and the intermediate results of such processing.

c. An output storage area holds the finished results of the processing operations until they can be released.

d. In addition to these data-related purposes, the primary storage section also contains a program storage area that holds the processing instructions.

The separate areas used for these general purposes are not fixed by built-in physical boundaries in the storage section. Rather, they can vary from one application to another. Thus, a specific physical space may store input data in one application, output results in another, and processing instructions in a third. The person writing the application instructions determines how the space will be used for each job.

2. The arithmetic-logic section. The arithmetic-logic and control sections together make up the central processing unit (CPU). All calculations are performed and all comparisons (decisions) are made in the arithmetic-logic section of the CPU. Once data are fed into primary storage from input devices, they are held and transferred as needed to the arithmetic logic section, where processing takes place. No processing occurs in primary storage. Intermediate results generated in arithmetic-logic unit are temporarily placed in a designated working storage area until needed at a later time. Data may thus move from primary storage to the arithmetic-logic unit and back again to storage many times before the processing is finished. Once completed, the final results are released to an input storage section and from there to an output device.

3. The Control section. By selecting, interpreting, and seeing to the execution of program instructions, the control section of the CPU maintains order and directs operations of the entire system. Although the control section doesn’t process data, it acts as a central nervous system for the data-manipulating components of the computer. At the beginning of processing: the first program instruction is selected and fed into the control section from the program storage area. There it is interpreted, and from there signals are sent to other components to execute the necessary action. Further program instructions are selected and executed, one after another, until the processing is completed.

4. Errors & Accidents

robot sent to disarm bomb goes wild in san francisco, read the head­line. Evidently, a hazardous-duty police robot started spinning out of con­trol when officers tried to get it to grasp a pipe bomb. Fortunately, it was shut off before any damage could be done. Most computer glitches are not so spectacular, although they can be almost as important.

In general, errors and accidents in computer systems may be classified as people errors, procedural errors, software errors, electromechanical problems, and "dirty data" problems.

• People errors: Recall that one part of a computer system is the people who manage it or run it. For instance, Brian McConnell of Roanoke, Virginia, found that he couldn't get past a bank's automated tele­phone system to talk to a real person. This was not the fault of the sys­tem so much as of the people at the bank. McConnell, president of a soft­ware firm, thereupon wrote a program that automatically phoned eight different numbers at the bank. People picking up the phone heard the recording, "This is an automated customer complaint. To hear a live com­plaint, press...”. Quite often, what may seem to be "the computer's fault" is human indifference or bad management.

• Procedural errors: Some spectacular computer failures have occurred because someone didn't follow procedures. Consider the 2 1/2-hour shut­down of Nasdaq, the nation's second largest stock market. Nasdaq is so automated that it likes to call itself "the stock market for the next 100 years." In July 1994, Nasdaq was closed down by an effort, ironically, to make the computer system more user-friendly. Technicians were phasing in new software, adding technical improvements a day at a time. A few days into this process, the technicians tried to add more features to the software, flooding the data-storage capability of the computer system. The result was a delay in opening the stock market that shortened the trading day.

• Software errors: We are forever hearing about "software glitches" or "software bugs." A software bug is an error in a program that causes it not to work properly.

An example of a somewhat small error was when a school employee in Newark, New Jersey, made a mistake in coding the school system's mas­ter scheduling program. When 1000 students and 90 teachers showed up for the start of school at Central High School, half the students had incom­plete or no schedules for classes. Some classrooms had no teachers while others had four instead of one. Especially with complex software, there are always bugs, even after the system has been thoroughly tested and debugged. However, there comes a point in the software development process where debugging must stop. That is, the probability of the bugs disrupting the system is considered to be low enough that it is not considered to be cost effective to find them and fix them.

• Electromechanical problems: Mechanical systems, such as printers, and electrical systems, such as circuit boards, don't always work. They may be incorrectly constructed, get dirty or overheated, wear out, or become damaged in some other way. Power failures (brownouts and blackouts) can shut a system down. Power surges can burn out equipment.

Modern systems, argues Yale University sociologist Charles Perrow, are made up of thousands of parts, all of which interrelate in ways that are impossible to anticipate. Because of that complexity, he says, what he calls "normal accidents" are inevitable. That is, it is almost certain that some combinations of minor failures will eventually amount to something cat­astrophic. Indeed, it was just such a collection of small failures that led to the blowing up of the Challenger space shuttle in 1986 and the near meltdown of the Three Mile Island nuclear-power plant in 1979. In the Digital Age, "normal accidents" will not be anomalies but are to be expected.

• "Dirty data" problems: When keyboarding a research paper, you undoubtedly make a few typing errors. So do all the data-entry people around the world who feed a continual stream of raw data into computer systems. A lot of problems are caused by this kind of "dirty data." Dirty data is data that is incomplete, outdated, or otherwise inaccurate.

A good reason for having a look at your records—credit, medical, school—is so that you can make any corrections to them before they cause you complications. As the president of a firm specializing in business intelligence writes, "Electronic databases, while a time-saving resource for the information seeker, can also act as catalysts, speeding up and magni­fying bad data."

An interesting source of "bad data," this same person notes, is the memos that every executive receives. He points out that "internal memos are rarely footnoted, or their 'facts' supported with proof. Very often the recipient of corporate information has to weigh its validity based on who said it," or on how badly he or she wants to believe it.

Дата добавления: 2015-11-14