Central Processing Unit
Central Processing Unit
Computers exist as a collection of interrelated components functioning together under the control of a central processor known as the central processing unit (CPU). The CPU is responsible for manipulating data and coordinating the activities of the computer's other physical components, including memory and peripherals. Instructions gathered from input interfaces are executed at the CPU, and the results delivered to output interfaces. The CPU, therefore, functions as the heart of the computer, facilitating all data processing activity.
The central processing unit is composed of several internal components needed to retrieve, store, and calculate data in a controlled fashion. Instructions enter the CPU from a computer's random access memory (RAM) through the bus . The bus is a grouping of wires that provide a physical medium for data transport between components. The instructions are decoded by the CPU's control unit, which interprets the data and sends control signals to the other components as appropriate. From here, instructions pass to the arithmetic logic unit (ALU), which performs calculations and other logical operations. The control unit and ALU depend on memory registers for the temporary storage of data and internal instructions. These registers, internal to the CPU, are similar to RAM but operate much faster and have far less storage capacity. They are used by the ALU to store calculated results until the end of an operation, and by the control unit to store instructions.
Computer instructions may be for data transfer, data manipulation, or program control. Data transfer instructions cause data to be moved between locations without affecting content, data manipulation instructions request arithmetic or logic operations from the ALU, and program control (branch) instructions facilitate decision operations. The control unit executes these instructions sequentially from consecutive memory locations. Each memory location is represented by a unique address, which permits a program counter to keep track of the last instruction executed. When the control unit retrieves an instruction, the program counter is incremented to reflect the next memory address. Unless the control unit is executing a branch instruction that alters this program counter value, that address will be the next instruction retrieved.
As noted earlier, the ALU performs arithmetic and logic operations. Basic arithmetic operations, like addition and subtraction, are performed by an arithmetic circuit; logic operations, such as AND, OR, and XOR (exclusive OR), are performed by a logic circuit. Like all components of the CPU, the ALU operates at the binary level. AND, OR, and XOR are examples of Boolean operations, whereby bits are compared to produce a logical (yes or no) result. A better understanding of ALU operations may be gained through the study of Boolean algebra .
Memory access is the slowest central processing operation; therefore memory registers are the most important components in determining the performance of a CPU. A register is a group of binary storage cells, or flip-flops, each capable of storing one bit of data. A CPU will often utilize large numbers of small registers because performance and capacity are inversely proportional, meaning that many small registers are faster than fewer larger registers. The smallest memory components are generally placed the closest to central processing components in order to optimize performance for the majority of processing operations.
A clock that sends repetitive pulses throughout the components of the CPU synchronizes all of these operations. Each clock pulse triggers an action—therefore, a CPU's performance can be measured by the frequency of clock pulses. The clock, however, must not exceed the performance of the registers or the CPU cannot function. The frequency of the clock is measured in Hertz (pulses per second).
Early CPUs were constructed from vacuum tubes , which required a great deal of energy and physical space compared to modern construction. The Electronic Numerical Integrator and Computer (ENIAC), which became operational in 1945 using more than 18,000 vacuum tubes, is largely regarded as the first electronic computer. The transistor was introduced in 1948, providing a smaller, faster, more efficient and reliable alternative to the vacuum tube. In 1956 the UNIVAC (Universal Automatic Computer) was completed, the first computer to incorporate a transistor-based CPU.
Development of the integrated circuit (IC), or computer chip, began in 1958 when Texas Instruments introduced a single piece of silicon containing multiple components. The integrated circuit provides the physical basis for today's microcomputers. In 1965 Gordon Moore made a prediction, now known as Moore's Law, that the number of transistors contained on a computer chip would double every year. In fact, the number of transistors integrated onto a single chip has doubled about every eighteen months over recent years. The first ICs had less than one hundred transistors, as opposed to the more than eight million transistors now common on a single chip. Continually improving methods in IC manufacturing have led to larger numbers of smaller components, which have in turn led to faster processing.
In 1967 Fairchild Semiconductor introduced an IC that contained all of the ALU functions, but required additional circuitry to provide register storage and data control. Intel Corporation introduced the first fully functioning microprocessor in 1971. The Intel 4004 was capable of four-bit arithmetic operations and was used in a number of handheld calculators.
The 4.77 MHz sixteen-bit Intel 8086 was introduced seven years later, becoming the first generation of the popular x86 series of microprocessors and the basis for the personal computer. This line of microprocessors, including the 80286 (286), 386, 486, and Pentium (586), has evolved to include a robust complement of digital components integrated within the same IC that contains the basic CPU. In 2001 Intel introduced the 32-bit Pentium IV with a clock speed of 1.5 GHz, or 1.5 billion pulses per second.
The integration of CPU and other computer functions on the same microprocessor chip has blurred the distinction between a computer and its CPU. It is not uncommon for computer users to refer to their entire system by the name of its CPU—a practice that is not unfounded since the architecture of a CPU largely determines every other peripheral the computer can support.
see also Intel Corporation; Microchip.
Jeffrey C. Wingard
Mano, M. Morris. Computer Systems Architecture. Englewood Cliffs, NJ: Prentice Hall, 1982.
Sclater, Neil. McGraw-Hill Electronics Dictionary. New York: McGraw Hill, 1997.