Introduction to microprocessors

The first microprocessor was introduced in 1970 by Intel (named 4004). It ran at the speed of 108KHz. Four years later, Intel created the 8080 running at just over 2 Mhz. This microprocessor was used on the world's firs personal computer, named Altair. Also at this time, IBM started researching for their microprocessor, called POWER (Performance Optimization With Enhanced RISC).

Motorola created their 68000 series of microprocessors in 1979. These were implemented later in the Macintosh computer by Apple.
Another significant role was held by Sun Microsystems when introducing Sparc (Scaleable Processor) in 1987. This generation of microprocessors used RISC (Reduced Instruction Set) making the processing operations faster.

Modern microprocessors contain millions of transistors concentrated on a small silicon square. Every transistor can go on or off, thus creating ones and zeros.

Regardless of their speed or physical form they have, the operation that most of the microprocessors do is to execute a sequence of stored instructions. Devices running according too the Von Neumann's architecure work in four steps: fetch, decode, execute and write.

Fetching involves getting an instruction from the program memory. An instruction is represented tipically by a sequence of numbers (eg. 101000101). This fetching operation forced reaserching for better memory technologies; the memory was slow enough to make the microprocessor's power unusable.
The next step is to decode the information, by breaking it into processable parts. Every part of the information goes to a portion of the CPU that can process it. This choice is made by considering the opcodes, and indicator posted before the information that signals what operation is needed (eg. addition).
Executing involves connecting various portions of the CPU toghether to serve the desired operation. Different schemes and connections between the parts form logical circuits that perform and act after Boolean rules. For example if a multiplication must be processed, the ALU (Arithmetic Logical Unit) will be linked with an Input and an Output address.
The final part, writing, takes place after the processing (execution) was completed. It involves writing the results to a a memory address given by a memory adressing scheme.

As far as the physical layout, microprocessors are usually connected to the motherboard through a socket using a PGA (Pin Grid Array).

Removing or adding a microprocessor can be a easy and safe operation if one knows what he is doing. It has a system called ZIF (Zero Insertion Force) that protects the processor's pins. As a caution, just remember to try not to touch the golden pins when working with a processor. After inserting it, it's wise to use a drop of thermal conductor paste betwen the microprocessor and a heatsink, to allow thermal transfer in the best conditions possible. Overheating due to heavy usage with poor ventilation will make the system instable and can harm it permantly.

Older versions of the Intel Pentium generation (II and some of III) used slots.

These type of processors usually had no ventilation, since their surface was large enough to cool itself. In the 1990s, when both slot and socket were used, some adapters also appeared on the market, like slot card with socket.

Subscribe now via RSS feed and get all the new tutorials