The Central Processing Unit
The central processing unit (CPU) is the heart of the computer of the computer system. Among other things, it configuration determines whether a computer is fast or slow in relation to other computers. The CPU is the most complex computer system component, responsible for directing most of the computer system activities based on the instructions provided. As one computer generation has evolved to the next, the physical size of the CPU has often become smaller, while its speed and capacity have increased tremendously. Indeed, these changes resulted in microcomputers that are small enough to fit on you desk or you lap. The CPU circuitry of a microcomputer-called a microprocessor-can fit on a chip about the size of your thumbnail, or even smaller.
The CPU has two main parts: (1) the control unit and (2) the arithmetic/logic unit. The parts of the CPU are usually connected by an electronic component referred to as a bus, which acts as an electronic highway between them. To temporarily store data and instructions, the CPU has special-purpose storage devices called registers.
Control unit
The control unit, a maze of complex electronic circuitry, is responsible for directing and coordinating most of the computer system activities. It does not execute instructions itself; it tells other parts of the computer system what to do. It determines the movement of electronic signals between main memory and the arithmetic/logic unit, as well as the control signals between the CPU and input/output devices.
Just as a car is useless without gas, a computer is not much good without software instructions. When we use software, we are working with high-level (human language-like) instructions that are to be carried out by the control unit. These instructions are converted by a language processor into a low-level form of instructions the computer can work with-machine language, the only language that the CPU can understand. In machine language, data and instructions are represented in binary form (0s and 1s), and each type of computer-microcomputer, minicomputer, or mainframe-responds to a unique version of machine language. Once the instructions have been converted into this form, they can be retrieved from main memory and interpreted by the control unit (sometimes referred to as decoding). According to each specific instruction, the control unit issues the necessary signals to other computer system components as needed to satisfy the processing requirements. This could involve, for example, directing that data be retrieved from a disk storage device, “telling” the printer to print the letter you just wrote, or simply directing the arithmetic/logic unit to add two numbers.
Arithmetic/Logic Unit (ALU)
Without the arithmetic/logic unit (ALU), the Kim program would not be able to do those financial forecasts for Sporting Life on the computer, and you would not be able to use Ticketron to find out if you and your friends can get five seats at the Lemon Bowl. In fact, without the ALU performs all the arithmetic and logical (comparison) functions-that is, it adds, subtract, multiplies, divides, and does comparisons. These comparisons, which are basically “less than”, “greater than”, and “equal to,” can be combined into several common expressions, such as “greater than or equal to.” The objective of most instructions that use comparisons is to determine which instruction should be executed next.
The ALU controls the speed of calculations and so receives a great deal of attention from computer engineers trying to meet the needs of the fast-paced business world. Older microcomputers’ speeds are usually measured in milliseconds-1 thousandth of a second. Never microcomputers’ speeds are measured in nanoseconds-1 billionth of a second-or picoseconds-1 trillionth of a second. If a nanosecond were equal to one minute, then a minute would be equal to 1,900 years!
Registers
Registers are special temporary storage locations within the CPU-some in the control unit and some in the ALU. Registers very quickly accept, store, and transfer data and instructions that are being used immediately. (Main memory holds data that will be used shortly; secondary storage-such as a diskette-holds data that will be used later). To execute an instruction, the control unit of the CPU retrieves it from main memory and places it into a register. The typical operations that take place in the processing of instructions are part of either the instruction cycle or the execution cycle.
The instruction cycle, or I-cycle, refers to the retrieval of an instruction from main memory and its subsequent decoding (the process of alerting the circuits in the CPU to perform the specified operation). The time it takes to go through the instruction cycle is referred to as I-time.
The execution cycle, or E-cycle, refers to the execution of the instruction and the subsequent storing of the result in a register. The time it takes to go through the execution cycle is referred to as E-time. The instruction cycle and the execution cycle together, as they apply to one instruction, are referred to as a machine cycle. The CPU has an internal clock that synchronizes all operations in the cycle. The speed is expressed in megahertz (MHz); 1MHz equals 1 million cycles per second. Generally, the faster the clock speed, the faster the computer can process information. Of course, for most users, the faster, the better… and the more expensive.
The number and types of registers in a CPU vary according to the CPU’s design. Their size (capacity) and number can dramatically affect the processing power of a computer system. In general, the “large” the register, the more bits can be processed at once. The size of a register is referred to as word-size. Some personal computers have general-purpose registers that hold only 8 bits; others hold 16 bits; newer microcomputers have 32-bit registers. Computers that handle a 32-bit word-size can process data twice as fast as those that handle a 16-bit word-size. The difference in processing power due to difference in register size can be illustrated by the difference between trying to put out a fire with a small drinking glass and with a 5-gallon bucket.
Bus
The term bus refers to an electrical pathway through which bits are transmitted between the various computer components. Depending on the design of a system, several types of buses may be present. For the user, the most important one is the data bus, which carries data throughout the CPU. the wider the data bus, the more data it can carry at one time and thus the greater the processing speed of the computer. The data bus in the Intel 8088 processor is 8 bits wide, meaning that it can carry 8 bits-or one character-at a time. In contrast, the data buses in the Motorola 68020 processor and the Intel 80386 processor are 32 bits wide-they can move four times more data through their data buses than the Intel 8088 bus can. Some supercomputers contain buses that are 128 bits wide..
日历