3.3.1. C P U (Microprocessor)

 Idi na Hrvatsku Stranicu  Back  Hardware  Next


Processor (CPU) is the most important electronic circuit computer program which takes command and orders and based on them, handles information. can consist of multiple hardware components, and the term 'micro' is used when all combined in one small enclosure. It contains millions of interconnected transistors on silicon substrates set in the plastic casing.

Time is of the order of execution microseconds, and the number of instructions and data sizes are basic binary parameters that define the quality of microprocessors. Microprocessors in general-purpose computer systems used are designed to work with data length of 4, 8, 16, 32 and 64 bits. The first is composed of a simple calculator, and the last one listed is more common practice in personal computers (Pentium).

Transfer data between the microprocessor circuitry within and among computers is very important part of the work, which is performed via BUS, stranded portable paths between individual circuits. Number of bus depends on the number of bits that describe the data that is necessary to be transferred. In essence, the data transmission is parallel.

Thus, there are two main types of bus:

            1.) Internal bus -- links within the microprocessor.
            2.) External bus -- connections from outside the microprocessor.

Are separated between the control circuits because often traffic to them does not happen at the same speed. Data exchange in the microprocessor performs many times faster than the circuits between the computers, and it is therefore necessary to separate. The transfer of data between internal and external bus care, special controller.

The development of microprocessors are related to the architecture with a common bus, as in the model in Figure 3.3.3, is not used anymore, but it is acceptable to display the principles of operation of the microprocessor.

 Block-scheme of CPU
Figure* 3.3.3 Basic microprocessor architecture with one bus. ( + / - )

Accumulator is most important register in the microprocessor. The tank is a temporary data storage. Data from the bus coming in and out of the battery through the mapped buffer B on the first ALU input. If you need to wait for the other data that will pass through the buffer B2 in the second ALU input which then performs an addition, logical operation, or transmit information to the registry and the rotation performs arithmetic operations. The result is then returned back to the bus, if necessary, re-supplied to the ALU for further processing or currently stored in one of the general purpose registers (R0 - RN) so that it is handy to the accumulator. The number of registers of general purpose processor depends on the design and are usually about 10-20, and one or more accumulators.

If the mathematical operation occurs overflows or other undesirable results, the status register (RS), which keeps a number of independent bits, changing the state of one of them, and points to the resulting change as the linesmen in football when FLAG signaling events. Insight into the content of the status register allows the programmer directing the flow of the program.

Thus, the general purpose registers store the data that are to be forwarded to accumulator and store the intermediate results of ALU. Besides the general-purpose registers, there are special-purpose registers are as follows:

Cooperation of all these components within a microprocessor can be best illustrated by a simple example stages adding two numbers:

  1. The program counter (PC) address set instructions.
  2. Instruction from that address is set in the instruction register (IR).
  3. Content instruction register (IR) is interpreted as an order to let the accumulator filled content found at that would be obtained in the next phase.
  4. Programme counter (PC) fill content is showing on the address information.
  5. Instruction register 'knows' that the data that is coming to be entered in data counter (DC) and a place where it is stored address operand.
  6. Fill the accumulator with the data that is stored at the address written down in the meter.
  7. The program counter (PC) 'fill' the address of the next instruction obtained from instruction register (IR), which requires that the contents of accumulator adding the content located at an address that has yet to be read.
  8. Obtaining the second operand according to the described procedure, and then the data is located on the inside of the bus.
  9. The data from the accumulator is reflected in the buffer with which it is connected to (B1) and data from the internal bus in the second buffer (B2).
  10. The control unit sends the AUL order to add these two binary data and the result is sent via the internal bus in one of the general purpose registers (R0 - RN).

This list describes a simplified microprocessor work, but it's also a good insight into the basics of how it works. The data are processed through buffers B0 be distributed outside the bus, and the control unit (CU) forwards the message to which the device to allow access to the external bus. From a simplified description of the microprocessor is obvious that all actions are performed in small increments, but the overall processing speed of these steps so large that their simplicity is not of importance.

Modern microprocessors have in your environment, besides the above circuits, various counters, memory, addressing and index containers and much more. Internal and external communication takes place over three separate bus:

            1.) Data bus ------ bidirectional
            2.) Address bus --- one way
            3.) Control bus --- one way

Of course, communication through three bus speeds up your system. Data that is on the data bus, address bus determines over the place that will be addressed, and the control bus transfer will be approved. All this was done on the model shown in the previous Figure alternately over a bus. With regard to increasing the efficiency of the design of microprocessors is first extended by adding her one ALU logical groups, as shown in Figure 3.5.3, with the reduction in the size of the elementary logic circuit and increasing the clock speed of microprocessors, and then the design has been extended in a way that is in the same casing logically connected multiple processor cores to simultaneously (in parallel) could execute multiple instructions, and even the modern versions have incorporated GPU as in Example chipset Intel Z77, Z87, Z97, z170, z270, z390 and z490.

Type of microprocessor is usually defined by the number of bits data bus and by how varied as 8 bit, 16 bit, 32 bit, 64 bit (most widely used) and others, or to a set of instructions (instruction set) used. If it comes to an extensive set of complex instructions, as well as the Intel x86 instruction set, it is a CISC (Complex Instruction Set Computing) architecture of processor, and if it is a small set of simple instructions it is a RISC (Reduced Instruction Set Computing) architecture of processor. According to this CISC processor and is able to perform complex tasks at once, and his logic is more complex and slower, until the contrary RISC processor with simpler logic and simple tasks per time unit able to execute multiple instructions than CISC processor. Frequently used combinations of these two approaches, the RISC microprocessor core type, while the remaining CISC, which means that a small number of complex instructions translates core microprocessor in a large number of simple instructions which is specific for a new generation of x86 microprocessors. CISC architecture prevalent in PCs, while the RISC architecture present in the PowerPC processors (which is once used by Mac) and simple processors for mobile devices, as ARM (Advanced RISC Machines) microprocessors, and Intel ATOM microprocessors, which does not use the x86 set of instructions, but simple micro-operations.

If it is on the simultaneous execution of multiple instructions for a superscalar microprocessor that are used for example Cray supercomputers, and the peculiarity of the multicore and / or multithreading processor (SMT - Simultaneous multithreading). If it is a special microprocessor dedicate - ASIC (Application-Specific Integrated Circuit), it is about a specific design for a particular purpose, such as DSP (Digital Signal Processor) processor with the task of converting analog video and audio signals to digital and vice versa, that uses SONAR or RADAR in his work. With some additional functionality microprocessors can operate as a separate small computers that control a process, but this is in Chapter 7.2 of microcontrollers.


Example I

What most computer users at least paying attention, but the reality is a computer and microprocessor biggest enemy, that is - TEMPERATURE. Heatsink is an indispensable condition of computer technology.

 Passive cooler  iStripper
Figure** 3.3.4 Passive heatsink (cooler). ( + / - )  

As an increasing number of transistors 'packed' into integrated circuits microprocessors operating temperature of the newer versions of microprocessors is increasingly growing. Excessive temperature rise may malfunction during operation and ultimately to burnout and damage the microprocessor. To avoid this microprocessor to cool down. Particular attention should be paid to the selection of cooling for the 'stronger' microprocessors; Active and passive cooling. The main feature of passive cooling is ribbed design (Figure 3.3.4a) and cooling without generating noise, but for modern processors, which is spending more than 100 W, are insufficient. Active cooling using substrate that is mounted on a microprocessor from which hot water (heat pipes), heat pipe radiator teaches a ribbed design and with ribs takes blowing with controlled fan. Cheaper Active cooling based on controlled cooling fins passive cooled by fans.

There are also special design with water cooling block that is mounted on a microprocessor and a heated water pump to drive a radiator and cools down, similar to a car engine cooling, or cooling performance with semiconductor plates (Peltier), if electricity is released through two different metals connected then metals react so that they will start 'switch' heat from one end to the other. Hence the name, and thermal pump. But it is a new burden on the power source.

Newer versions of microprocessors use the mechanism of automatic lowering the clock speed when the processor is not loaded as the warming effect as much as possible. In order to heat (dissipation) of the body with the microprocessor to better convey the cooling surface is used thermal paste in a thin layer is applied on the metal casing before placing the microprocessor heat sink, in order to reduce the effect of joining uneven surfaces. In this way a greater or lesser extent enhances the exhaust heat from the microprocessor depending on the type and characteristics of the paste. Although the microprocessor in the computer cools in a special way, a lot of other very heated components within the enclosure that require good air cooling. Therefore, the type of casing and ventilation enclosures should be given special attention.

 Features of industrial computer

Special event are rugged and industrial computers, as in the Figure above this paragraph and shown in Figures 3.3.4b and 3.3.4c, which have to work in dusty conditions, and the cooling fan as a cooling component is not a solution. Then the whole case is designed as a cooler. Special attention was dedicated to the software that should not stop working without warning, and the work of software supervises special counter (Watch Dog Timer) as at the microcontroller. Pay attention to the operating temperature range shown in the Figure above.


 Back
 Search
 Next

Citing of this page:
Radic, Drago. " Informatics Alphabet " Split-Croatia.
{Date of access}; https://informatics.buzdo.com/file.
Copyright © by Drago Radic. All rights reserved. | Disclaimer
 Content - Home
 Content  Informatics Alphabet