System unit. General information The system unit is the main unit, inside which the most important components are installed. Devices located. Presentation on the topic "system unit" Presentation on the topic

Slide 3

CPU

• During operation, the processor reads a sequence of instructions contained in memory and executes them. This sequence of commands is called a program and represents the algorithm for the useful operation of the processor. The order of reading commands changes if the processor reads a jump command - then the address of the next command may be different. Another example of a process change would be when a stop command is received or switches to hardware interrupt mode.
• The processor is the main chip of the computer, its “brain”. It executes program code located in memory and controls the operation of all computer devices. The faster the processor speed, the faster the computer will perform. The processor has special cells called registers. It is in the registers that the commands that are executed by the processor are placed, as well as the data that the commands operate on. The processor's job is to select instructions and data from memory in a specific sequence for subsequent execution.

Slide 4

Internal memory and its characteristics

• Designed for temporary storage of executed programs and data processed by these programs. This is volatile memory. Physically implemented in RAM modules (random access memory devices) of various types. When the power is turned off, all information in RAM disappears.
• The amount of information stored in RAM ranges from 32 to 512 MB or more. Entering information into memory and retrieving it is carried out at addresses. Each byte of the OP has its own individual address (sequence number). Address is a number that identifies memory cells (registers). The OP consists of a large number of cells, each of which stores a certain amount of information. The OP is directly connected to the processor. The capabilities of a PC largely depend on the volume of OP.

Slide 5

Cache memory

Cache memory is a very fast, small-volume storage device that is used when exchanging data between the processor and RAM to compensate for the difference in the speed of information processing between the processor and the somewhat slower RAM. The cache memory is controlled by a special controller, which, by analyzing the program being executed, tries to predict what data and commands the processor will most likely need in the near future, and pumps them into the cache memory. English Cache - secret warehouse

Slide 6

Drives and controllers

• A disk drive is an essential peripheral device. As a peripheral device, it requires an interface card. This card is called a drive controller card. In the most general case, this card does not create any problems. If they do happen, you need to know how to detect and identify them.
• All it takes is a little dirt to cause failure on both the floppy disk and the drive. Disk drives require speed adjustments, head alignment and cleaning. Adjusting the rotation speed and cleaning the heads is relatively simple. Adjusting the heads requires special equipment and is not always cost-effective, but this will be discussed later.
• Besides alignment, there is also the issue of device compatibility. There are currently two types of drives: 3.5" and 5.25"; There are also “floptical” devices that allow you to store about 100 M of information on one disk with a diameter of 3.5"
• If you're building a new computer, you'll most likely need one 3.5" floppy drive, since most new software is distributed on floppy disks. The article on hard drives goes into detail about servicing floppy drives.

Slide 7

System unit ports

• A port is usually a connection (physical or logical) through which data is received and sent in computers.
• The most commonly called port is:
• A hardware port is a specialized connector on a computer designed to connect equipment of a certain type. See: LPT port, serial port, USB port, Game port.
• I/O Port - Used in microprocessors (e.g. Intel) and microcontrollers (e.g. PIC, AVR) when communicating with hardware. An I/O port is associated with a device and allows programs to access it to exchange data.
• Network port is a parameter of the TCP and UDP protocols.

Slide 8

View all slides

General information The system unit is the main unit, inside which the most important components are installed. Devices located inside the system unit are called internal, and devices connected to it from the outside are called external. External additional devices designed for input, output and long-term storage of data are also called peripherals. In appearance, system units differ in the shape of the case. Personal computer cases are produced in horizontal (desktop) and vertical (tower) versions. Cases that have a vertical design are distinguished by size: full-size (big tower), mid-size (midi tower) and small-size (mini tower). Among the cases that have a horizontal design, there are flat and especially flat (slim). Personal computer cases are supplied with a power supply and, thus, the power of the power supply is also one of the case parameters. For mass models, the power supply W is sufficient.

Motherboard Motherboard is the main board of a personal computer. It contains: a processor - the main chip that performs most mathematical and logical operations; a processor - a microprocessor set (chipset) - a set of chips that controls the operation of the internal devices of the computer and determines the main functionality of the motherboard; a microprocessor set - random access memory (random access memory, RAM) - a set of chips designed for temporary storage of data when the computer is turned on; RAM buses are sets of conductors through which signals are exchanged between the internal devices of the computer;

Hard drive The hard drive is the main device for long-term storage of large amounts of data and programs. In fact, it is not one disk, but a group of main disks that are magnetically coated and spin at high speed. Thus, this “disk” does not have two surfaces. Above each surface is a head designed for reading and writing data. At high disk rotation speeds (r/s), an aerodynamic cushion is formed in the gap between the head and the surface, and the head hovers above the magnetic surface at a height of several thousandths of a millimeter. When the current flowing through the head changes, the intensity of the dynamic magnetic field in the gap changes, which causes changes in the stationary magnetic field of the ferromagnetic particles that form the coating of the disk. This is how data is written to a magnetic disk. The reading operation occurs in reverse order. Magnetized coating particles flying at high speed near the head induce a self-induction emf in it. The electromagnetic signals generated in this case are amplified and transmitted for processing. The operation of the hard disk is controlled by a special hardware-logical device, the hard disk controller. In the past, it was a separate daughterboard that was connected to one of the free slots on the motherboard. Currently, the functions of disk controllers are partially integrated into the hard drive itself, and partially performed by microcircuits included in the microprocessor kit (chipset).

Floppy disk drive Information on a hard drive can be stored for years, but sometimes you need to transfer it from one computer to another. Despite its name, a hard drive is a very fragile device, sensitive to overloads, shocks and shocks. Theoretically, it is possible to transfer information from one workplace to another by moving a hard drive, and in some cases this is done, but still this technique is considered low-tech, since it requires special care and certain qualifications. To quickly transfer small amounts of information, so-called flexible magnetic disks (floppy disks) are used, which are inserted into a special drive. The drive's receiving hole is located on the front panel of the system unit. The correct direction for feeding the floppy disk is indicated by an arrow on the plastic casing. The main parameters of floppy disks are: technological size (measured in inches), recording density (measured in multiples of units) and total capacity. 3.5-inch floppy disks have been produced since 1980. The single-sided regular-density disc had a capacity of 180 KB, double-sided 360 KB, and double-sided double-density 720 KB. Nowadays, 3.5-inch high-density disks are considered standard. They have a capacity of 1440 KB (1.4 MB) and are marked with the letters HD (high density).

CD-ROM drive Over the years, floppy drives with a diameter of 5.25 inches were no longer included in the basic configuration of personal computers, but instead the installation of a CD-ROM drive with the same external dimensions began to be considered standard. The abbreviation CD-ROM (Compact Disc Read-Only Memory) is translated into Russian as a permanent storage device based on a compact disc. The operating principle of this device is to read numerical data using a laser beam reflected from the surface of the disk. Digital recording on a CD differs from recording on magnetic disks in its very high density, and a standard CD can store approximately 650 MB of data. Large amounts of data are typical for multimedia information (graphics, music, video), so CD-ROM drives are classified as multimedia hardware. Software products distributed on CDs are called multimedia publications. Today, multimedia publications are gaining an increasingly stronger place among other traditional types of publications. For example, there are books, albums, encyclopedias and even periodicals (electronic magazines) published on CD-ROM. The main disadvantage of standard CD-ROM drives is the inability to write data, but in parallel with them, there are also CD-ROM burners today, CD-RW drives. Special blanks are used for recording. Some of them allow only one write (after burning, the disc turns into a regular read-only CD-ROM), others allow you to erase previously recorded information and write again. The main parameter of CD-ROM drives is the data reading speed. It is measured in multiples. The unit of measurement is the reading speed of music CDs, which in terms of data is 150 KB/s.

Video card Video card Together with the monitor, the video card forms the video subsystem of a personal computer. The video card has not always been a PC component. At the dawn of the development of personal computing technology, in the general area of ​​RAM there was a small dedicated screen memory area into which the processor entered image data. A special screen controller read data on the brightness of individual points of the screen from the memory cells of this area and, in accordance with them, controlled the scanning of the horizontal beam of the monitor's electron gun. With the transition from black-and-white monitors to color ones and with an increase in screen resolution (the number of pixels vertically and horizontally), the video memory area became insufficient to store graphic data, and the processor could no longer cope with constructing and updating the image. It was then that all operations related to screen control were separated into a separate block, called the video adapter. Physically, the video adapter is designed as a separate daughter card, which is inserted into one of the slots on the motherboard and is called a video card. The video adapter took over the functions of the video controller, video processor and video memory. During the existence of personal computers, several video adapter standards have changed: MDA (monochrome)] CGA (4 colors), EGA (16 colors); VGA (256 colors). Currently, SVGA video adapters are used, providing optional playback of up to 16.7 million colors with the ability to arbitrarily select screen resolution from a standard range of values ​​(640x480, 800x600, 1024x768, 1152x864; 1280x1024 pixels and more). Screen resolution is one of the most important parameters of the video subsystem. The higher it is, the more information can be displayed on the screen, but the less. the size of each individual point and, accordingly, the smaller the visible size of the image elements.

Sound card The sound card was one of the most recent improvements in the personal computer. It is installed in one of the connectors of the motherboard in the form of a daughter card and performs computational operations related to the processing of sound, speech, and music. The sound is played through external sound speakers connected to the sound card output. A special connector allows you to send an audio signal to an external amplifier. There is also a microphone connector, which allows you to record speech or music and save it on your hard drive for later processing and use. The main parameter of a sound card is the bit depth, which determines the number of bits used when converting signals from analog to digital form and vice versa. The higher the bit depth, the smaller the error associated with digitization, the higher the sound quality. The minimum requirement today is 16 bits, and 32-bit and 64-bit devices are the most common. In the field of audio reproduction, the most difficult situation is with standardization. In the absence of uniform centralized standards, devices compatible with the SoundBlaster device, which is trademarked by Creative Labs, have become the de facto standard. Recently, audio processing has been viewed as a relatively simple operation, which, due to the increased power of the processor, can be entrusted to it. In the absence of increased requirements for sound quality, integrated sound systems can be used, in which sound processing functions are performed by the central processor and motherboard chips. In this case, speakers or other audio playback devices are connected to sockets installed directly on the motherboard.

Random Access Memory Random Access Memory (RAM) is an array of crystalline cells capable of storing data. There are many different types of RAM, but from the point of view of the physical principle of operation, they distinguish between dynamic memory (DRAM) and static memory (SRAM). dynamic memory static memory Each memory cell has its own address, which is expressed as a number. In most modern processors, the address size limit is usually 32 bits, which means that there can be a total of 232 independent addresses. One addressable cell contains eight binary cells that can store 8 bits, that is, one byte of data. In modern computers, it is possible to directly address a memory field of 232 bytes = 4 GB. The maximum size of the RAM field installed in a computer is determined by the microprocessor set (chipset) of the motherboard and usually cannot exceed several GB. The minimum amount of memory determined by the operating system requirements for modern computers is 128 MB.

Today, the typical RAM size is 256 MB, but the upward trend continues. RAM in a computer is located on standard panels called modules. RAM modules are inserted into the corresponding slots on the motherboard. The main characteristics of RAM modules are memory capacity and data transfer speed. Today, the most common modules are MB in size. The data transfer rate determines the maximum memory bandwidth (in MB/s or GB/s) in the optimal access mode. This takes into account memory access time, bus width and additional capabilities, such as transmitting multiple signals in one clock cycle. Modules of the same volume can have different speed characteristics. Sometimes access time is used as a defining characteristic of memory. It is measured in billionths of a second (nanoseconds). For modern memory modules, this value can be 5 ns, and for particularly fast memory, used mainly in video cards, it can be reduced to 2-3 ns. RAM

Dynamic memory Dynamic memory (DRAM) cells can be thought of as microcapacitors that can store charge on their plates. This is the most common and economically available type of memory. The disadvantages of this type are associated, firstly, with the fact that both when charging and discharging capacitors, transient processes are inevitable, that is, data recording occurs relatively slowly. The second important drawback is related to the fact that cell charges tend to dissipate in space, and very quickly. If the RAM is not constantly “recharged,” data loss occurs within a few hundredths of a second. To combat this phenomenon, the computer undergoes constant regeneration (refreshing, recharging) of RAM cells. Regeneration occurs several tens of times per second and causes wasteful consumption of computing system resources. Dynamic memory chips are used as the main RAM of a computer.

Static memory (SRAM) Static memory cells (SRAM) can be thought of as microelectronic flip-flops consisting of several transistors. The trigger stores not the charge, but the state (on/off), so this type of memory provides higher performance, although it is technologically more complex and, accordingly, more expensive. Static memory chips are used as auxiliary memory (called cache memory) designed to optimize processor performance.

Processor The processor is the main computer chip in which all calculations are performed. Structurally, the processor consists of cells similar to RAM cells, but in these cells data can not only be stored, but also changed. The internal cells of the processor are called registers. It is also important to note that data placed in some registers is not considered as data, but as instructions that control the processing of data in other registers. Among the processor registers there are those that, depending on their content, are capable of modifying the execution of commands. Thus, by controlling the sending of data to different registers of the processor, you can control the processing of data. This is what program execution is based on.

ROM chip and BIOS system When the computer is turned on, there is nothing in its RAM, neither data nor programs, since RAM cannot store anything without recharging the cells for more than hundredths of a second, but the processor needs commands, including at the first moment after turning on . Therefore, immediately after switching on, the start address is set on the processor address bus. This happens in hardware, without the participation of programs (always the same). The processor addresses the set address for its first command and then begins to work according to the programs. This source address cannot point to RAM, which does not yet have anything in it. It indicates a different type of memory, read-only memory (ROM). The ROM chip is capable of storing information for a long time, even when the computer is turned off. Programs located in ROM are called “hardwired”; they are written there at the stage of chip manufacturing

System unit computer

Motherboard

· Motherboard - the main board of a personal computer. It contains 1:

· processor - the main chip that performs most mathematical and logical operations;

· microprocessor kit (chipset) - a set of microcircuits that control the operation of the internal devices of the computer and determine the basic functionality of the motherboard;

· buses - sets of conductors through which signals are exchanged between the internal devices of the computer;

· random access memory (random access memory, RAM) - a set of chips designed for temporary storage of data when the computer is turned on;

· ROM (read only memory) is a chip designed for long-term storage of data, including when the computer is turned off;

· connectors for connecting additional devices (slots).

We will consider the devices included in the motherboard separately.

Hard drive

Hard drive -- the main device for long-term storage of large amounts of data and programs. In fact, this is not one disk, but a group of coaxial disks that have a magnetic coating and rotate at high speed. Thus, this “disk” does not have two surfaces.

Above each surface is a head designed for reading and writing data. At high disk rotation speeds (90-250 rps), an aerodynamic cushion is formed in the gap between the head and the surface, and the head hovers above the magnetic surface at a height of several thousandths of a millimeter. When the current flowing through the head changes, the intensity of the dynamic magnetic field in the gap changes, which causes changes in the stationary magnetic field of the ferromagnetic particles that form the coating of the disk. This is how data is written to a magnetic disk.

The reading operation occurs in reverse order. Magnetized coating particles flying at high speed near the head induce a self-induction emf in it. The electromagnetic signals generated in this case are amplified and transmitted for processing.

The hard drive is controlled by a special hardware-logical device -- hard drive controller. In the past it was a separate daughter board, which was connected to one of the free slots on the motherboard. Currently, the functions of disk controllers are partially integrated into the hard drive itself, and partially performed by microcircuits included in the microprocessor kit (chipset).

Floppy drive

Information on a hard drive can be stored for years, but sometimes it needs to be transferred from one computer to another. Despite its name, a hard drive is a very fragile device, sensitive to overloads, shocks and shocks. Theoretically, it is possible to transfer information from one workplace to another by moving a hard drive, and in some cases this is done, but still this technique is considered low-tech, since it requires special care and certain qualifications.

To quickly transfer small amounts of information, so-called floppy disks(floppy disks) that are inserted into a special drive -- drive. The drive's receiving hole is located on the front panel of the system unit. The correct direction for feeding the floppy disk is indicated by an arrow on the plastic casing.

The main parameters of floppy disks are: technological size (measured in inches), recording density (measured in multiples of units) and total capacity.

First computer IBM PC(the founder of the platform) was released in 1981. It could be connected to an external drive using single-sided floppy disks with a diameter of 5.25 inches. The disk capacity was 160 KB. The following year, similar double-sided disks with a capacity of 320 KB appeared. Beginning in 1984, 5.25-inch high-density (1.2 MB) floppy disks were produced. These days, 5.25-inch drives are not used, so the production and use of 5.25-inch drives has virtually ceased since the mid-90s.

3.5-inch floppy disks have been produced since 1980. Single sided disc regular density had a capacity of 180 KB, double-sided - 360 KB, and double-sided double density -- 720 KB. Nowadays, 3.5-inch wheels are considered standard. high density. They have a capacity of 1440 KB (1.4 MB) and are marked with the letters HD (high density - high density).

CD-ROM drive

In the period 1994-1995, the basic configuration of personal computers no longer included floppy drives with a diameter of 5.25 inches, but instead the installation of a floppy drive began to be considered standard CD-ROM having the same external dimensions.

Abbreviation CD-ROM (Compact Disc Read-Only Memory) translated into Russian as CD-ROM based read-only storage device. The operating principle of this device is to read numerical data using a laser beam reflected from the surface of the disk. Digital recording on a CD differs from recording on magnetic disks in its very high density, and a standard CD can store approximately 650 MB of data.

Large volumes of data are typical for multimedia information(graphics, music, video), so disk drives CD-ROM classified as multimedia hardware. Software products distributed on CD-ROMs are called multimedia publications. Today, multimedia publications are gaining an increasingly stronger place among other traditional types of publications. For example, there are books, albums, encyclopedias and even periodicals (electronic magazines) published on CD-ROM.

The main disadvantage of standard disk drives CD-ROM is the impossibility of recording data, but in parallel with them today there are also CD recording devices - disk drives CD-RW. Special blanks are used for recording. Some of them can only be written once (after burning, the disc turns into a regular CD CD-ROM read-only), others allow you to erase previously recorded information and record again.

The main parameter of disk drives CD-ROM is the data reading speed. It is measured in multiples. The unit of measurement is the reading speed of music CDs, which in terms of data is 150 KB/s.

Video card (video adapter)

Together with the monitor video card forms video subsystem personal computer. The video card has not always been a PC component. At the dawn of the development of personal computing technology, in the general area of ​​RAM there was a small dedicated screen memory area, into which the processor entered image data. Special screen controller read data on the brightness of individual points on the screen from memory cells of this area and, in accordance with them, controlled the scanning of the horizontal beam of the monitor’s electron gun.

With the transition from black and white monitors to color and with increasing screen resolution(number of points vertically and horizontally) the video memory area became insufficient to store graphic data, and the processor could no longer cope with constructing and updating the image. Then all operations related to screen control were separated into a separate block, called video adapter Physically, the video adapter is designed as a separate daughter board, which is inserted into one of the slots on the motherboard and is called video card. The video adapter took over the functions video controller, video processor And video memory.

During the existence of personal computers, several video adapter standards have changed: MDA (monochrome)] CGA (4 colors), EGA (16 flowers); VGA (256 flowers). Currently used video adapters SVGA providing optional reproduction of up to 16.7 million colors with the ability to randomly select screen resolution from a standard range of values ​​(640x480, 800x600, 1024x768, 1152x864; 1280x1024 pixels and beyond).

Screen resolution is one of the most important parameters of the video subsystem. The higher it is, the more information can be displayed on the screen, but the smaller the size of each individual dot and, accordingly, the smaller the visible size of the image elements.

The video subsystem of any computer consists of two parts - the video adapter, which is inserted into the expansion connector on the motherboard, and the display, which is connected to the video adapter.

The video adapter can be designed as a separate card that is inserted into the computer's expansion slot, or it can be located directly on the computer's motherboard.

The video adapter includes video memory, which stores the image currently displayed on the display screen, a read-only memory device, which stores the font sets displayed by the video adapter in text and graphics modes, as well as BIOS functions for working with the video adapter. In addition, the video adapter contains a complex control device that ensures data exchange with the computer, image formation and some other actions.

Video adapters can operate in various text and graphic modes, differing in resolution, number of displayed colors and some other characteristics.

The video adapter itself does not display data. To do this, you need to connect a display to the video adapter. The image generated by the computer is generated by the video adapter and transmitted to the display for presentation to the end user.

The video adapter is designed to store video information and display it on the monitor screen. It directly controls the monitor, as well as the process of displaying information on the screen by changing the horizontal and vertical scanning signals of the CRT monitor, the brightness of image elements and color mixing parameters. The main components of a modern video adapter are the video controller itself (usually a custom LSI - ASIC), video BIOS, video memory, a special digital-to-analog converter RAMDAC (Random Access Memory Digital to Analog Converter), a crystal oscillator (one or more) and interface chips with the system bus (ISA, VLB, PCI, AGP or other). An important element of the video subsystem is its own memory. For this purpose, the memory of the video adapter is used, which is often also called video memory, or frame buffer, or part of the PC RAM (in UMA shared memory architecture).

All modern video subsystems can operate in one of two main video modes: text or graphic. In text mode, the monitor screen is divided into separate character positions, each of which can display only one character at a time. To convert character codes stored in the video memory of the adapter into bitmaps on the screen, a so-called character generator is used, which is usually a ROM where character images are stored, “laid out” on lines. When receiving a character code, the character generator generates a corresponding binary code at its output, which is then converted into a video signal. Text mode in modern operating systems is used only at the initial boot stage.

Video memory

There are two questions: how many, and what type? As for the volume, it is at least two megabytes. Moreover, the amount of memory required is directly related to the resolution with which you plan to work and the depth of color representation.

Permission. The number of pixels represented by bits in video memory, or the addressable resolution. Video memory can be organized by the ratio of pixels (bits) along the x-axis (pixels on a row) to the number of pixels along the y-axis (columns) and to the size of memory allocated to represent color depth. Standard VGA video memory is 640 pixels by 480 pixels and typically has a color depth of 8 bits. The higher the resolution, the more detailed the image, and the more information about it needs to be stored. But not all stored information can be displayed on the display.

Pixel. A combined term for a picture element that is the smallest element of a monitor screen. Another name is pel.

The image on the screen is made up of hundreds of thousands of pixels combined to form an image. A pixel is the smallest segment of a raster line that is discretely controlled by the system that produces the image. On the other hand, it is a coordinate used to determine the horizontal spatial position of a pixel within an image. Pixels on a monitor are luminous dots of bright phosphorus that are the minimal element of a digital image. The pixel size cannot be smaller than the dot that the monitor can form. On a color monitor, dots consist of groups of triads. Triads are formed by three different phosphorus: red, green and blue. The phosphors are located along the sides of each other. Pixels may vary in size and shape depending on the monitor and graphics mode. The number of dots on the screen is determined by the physical ratio of the width to the height of the tube.

And here's why:

Some technical details.

Before becoming an image on the monitor, binary digital data is processed by the central processor, then sent through the data bus to the video adapter, where it is processed and converted into analog data, and after that it is sent to the monitor and forms an image. First, the digital data from the bus enters the video processor, where it begins to be processed. After this, the processed digital data is sent to video memory, where an image of the image is created, which should be displayed on the display.

Then, still in digital format, the data that makes up the image is transferred to the RAMDAC, where it is converted to analog and then transferred to a monitor, which displays the desired image.

Thus, almost along the entire path of digital data, various transformation, compression and storage operations are performed on it. By optimizing these operations, you can improve the performance of the entire video subsystem. Only the last leg of the path, from RAMDAC to the monitor, when the data is in analog form, cannot be optimized.

Let's take a closer look at the stages of data flow from the system's central processor to the monitor.

1. The speed of data exchange between the CPU and the graphics processor directly depends on the frequency at which the bus through which the data is transmitted operates. The bus operating frequency depends on the motherboard chipset. For video adapters, the PCI and AGP buses are optimal in terms of speed. The higher the operating frequency of the bus, the faster the data from the system's central processor reaches the graphics processor of the video adapter.

2. The key point that affects the performance of the video subsystem, regardless of the specific functions of various graphics processors, is the transfer of digital data processed by the graphics processor to video memory, and from there to RAMDAC. The bottleneck of any video card is the video memory, which continuously services the two main devices of the video adapter: the graphics processor and RAMDAC, which are always overloaded with work. At any time when changes occur on the monitor screen (sometimes they occur continuously, for example, moving the mouse pointer, blinking the cursor in the editor, etc.), the GPU accesses the video memory. At the same time, RAMDAC must continuously read data from video memory so that the image does not disappear from the monitor screen. Therefore, to increase video memory performance, manufacturers use various technical solutions. For example, they use different types of memory with improved properties and advanced capabilities, for example, VRAM, WRAM, MDRAM, SGRAM, or increase the width of the data bus on which the GPU or RAMDAC communicates with the video memory, using 32-bit, 64-bit or 128 - bit video bus.

One option is to use dual-port video memory.

Those. The GPU reads from or writes to video memory through one port, and the RAMDAC reads data from video memory using a second independent port. As a result, the GPU no longer has to wait for the RAMDAC to complete its VRAM operations, and conversely, the RAMDAC no longer has to wait for the GPU to complete its VRAM operations.

Another method for increasing performance is to increase the width of the bus through which the GPU and RAMDAC communicate with the video memory.

But the most common method today for optimizing the operation of video adapters is the use of an increased clock frequency at which the graphics processor, video memory and RAMDAC operate, which allows increasing the speed of information exchange between board components.

RAMDAC operating principles and parameters

RAMDAC has two operating modes. In the first mode, the chipset operates with color gamut or palette data (palletized data). In this mode, 8-bit data is converted to RGB colors. Each of the 256 possible color values ​​corresponds to a position in the color palette, which is located in the DAC (Digital to Analog Converter). The color palette is generated and stored in RAM (random access memory) - hence the name RAMDAC - and can be loaded with any color combination. Each time a new pixel is sent to the DAC for display on the screen, the value of the data transferred is used as a pointer to the position in the palette, the information from the palette is used as the color value for the DAC. The palette stored in RAM has 256 positions, each of which stores 24 bits of color data, 8 bits for each of the three primary colors Red, Green and Blue. RAM capacity corresponds to the value 256 x 24 = 6144 bits or 768 bytes. For RAM, standard memory is used, made using DRAM technology and integrated together with the graphics controller and DAC into one chip, in other words, into one silicon (silicon).

By the way, the technology for including RAM for the DAC in the graphics chipset has nothing to do with the so-called Embedded RAM. The latter is used as local memory, also called a frame buffer.

In the second mode, RAMDAC operates on color data. In this mode (with 16, 24 or 32 bit color representation), the data is RGB color. For example, with a 16-bit color representation, 5 bits define Red, 6 bits Green, and 5 bits Blue. For green, more bits are used because the human eye is more sensitive to green. With 24 or 32 bit color representation, 8 bits of data are used for each color. In this mode, color data is transferred directly to the DAC without the use of RAM, i.e. no loaded palettes are used and data is transferred directly from video memory. Since RAM is not involved, there is no 205 MHz limit for the frequency at which the DAC operates. The only limitation is the maximum possible speed of the DAC.

Selecting the RAMDAC operating mode occurs as follows: the Windows95/98/NT operating system or application reports the required mode to the video adapter driver, which switches the RAMDAC to one or another operating mode. The Display Control Panel utility in Windows allows you to choose between 8, 16 or 24/32 bit color representation. This is the way Windows selects the operating mode of the RAMDAC. An application that runs in full screen can set any mode it requires, the main thing is that this mode is supported by the video adapter.

The operating system or driver makes a request to determine the resolution, color depth, and refresh rate of the screen. The driver can either implement the response received or return a message indicating that the requested mode is not supported or is not possible. In this case, the operating system or application should try to request a different video mode setting.

The choice of RAMDAC operating mode has nothing to do with the type of video memory used.

The choice of mode in which the RAMDAC operates depends on the number of possible colors. DAC has a capacity of 8*8*8 bits, i.e. 8 bits for each RGB color, which corresponds to the ability to display 16,777,216 (16M) colors. With an 8-bit color representation, 256 of the 16 million possible colors can be used for the palette. When using color gamut (palette) data, only 256 colors are active and can be displayed on the screen at any randomly selected point in time. However, the palette can be changed by the application at any time. With an 8-bit color representation depth, each application is responsible for loading the palette. With 16-bit color, there is a fixed set of colors and any of the 65536 (64K) colors available can be used for display. At 24-bit or 32-bit color, the DAC can display any of 16 million (16M) possible colors.

Every user can notice that with 8-bit color, any graphic image does not look as good as with 16-bit color representation. However, most users cannot tell the difference when viewing a well-made graphic in 16-bit and 32-bit color modes. The phrase " well done graphic image" means rasterization(dithering) is the process of mixing two adjacent colors to obtain a third while simultaneously ensuring smooth transitions between image elements. As a result of the use of rasterization technology, images are obtained that look almost the same in modes with different color depths.

A 16-bit color representation requires twice as much memory as an 8-bit color representation, and a 32-bit color representation requires twice as much memory as a 16-bit color representation. Due to the fact that graphics adapters have limited amounts of memory, saving this resource becomes one of the priorities. Additionally, displaying 32-bit data often takes longer than displaying 16-bit data. And this already relates to the problem of performance, which should not be forgotten either. This is why the average user should use 16-bit color representation in Windows95/98/NT.

The user or application selects the color presentation mode that is most convenient for them. Word processors, spreadsheets, and 2D games can run perfectly in 8-bit color mode. Videos, 3D games and 3D applications typically use 16-bit color as a trade-off between image quality and performance. When using programs for viewing high-quality photos, editing them, as well as applications for creating graphics, it is best to use 24/32-bit color representation.

How can you find out what mode the RAMDAC is operating in? If you are using Windows, then you have the opportunity to select the color depth between 8, 16 or 24/32 bit modes. In 8-bit mode, a palette is used, i.e. RAMDAC runs at 205 MHz, in all other modes, with a different color depth, no palette is used and RAMDAC runs at 220 MHz. If an application running in full-screen mode is launched (for example, most games run in this mode), then the application itself determines in which mode the RAMDAC will operate. Sometimes the application, by selecting an operating mode, communicates this information to the user. But in most cases this does not happen.

The user can find out which mode the RAMDAC is operating in by doing the following: Find a surface that has a smooth transition from one color to another (like the sky above your head). If the transition from one color to another looks like it consists of alternating dots of very different colors, then your application is running in 8-bit color mode. Otherwise, i.e. if the transition from one color to another is really smooth, your application is running at a different color depth. At the same time, it is worth recalling once again that the average user cannot determine with confidence what color depth he is dealing with, 16 or 24/32 bits.

It is quite simple to make sure that the stated RAMDAC speed values ​​are true. If you know what resolution you are working at, for example 1024x768, and at what refresh rate the image is refreshed, for example 75 Hz, then you can find out what the speed of the DAC is. A speed of 220 MHz is quite enough to display in 1280x1024 at 85 Hz and 1600x1200 at 75 Hz modes. For 1600x1200 mode at 85 Hz, a speed of 250 MHz is required. It is known that according to European standards, a screen refresh rate of 85 Hz must be supported in all resolutions, but only a few models of modern monitors can operate in 1600x1200 mode at 85 Hz.

Let us recall the known facts: if the screen refresh rate is too low, the user will notice flickering of the image, which can result in damage to vision. A screen refresh rate of 75 Hz is already fast enough for the human eye to notice flicker. Therefore, it makes much more sense to focus on the refresh rate values ​​rather than the DAC speed, especially since these values ​​are interrelated.

Graphics accelerators(accelerators) - specialized graphics coprocessors that increase the efficiency of the video system. Their use frees the central processor from a large amount of operations with video data, since the accelerators independently calculate which pixels to display on the screen and what their colors are.

Video accelerators

The image that we see on the monitor screen is the contents of the video memory output by a special digital-to-analog converter RAMDAC (Random Access Memory Digital to Analog Converter) and a scanner. This content can be changed by both the central processor and the graphics card's graphics processor - a two-dimensional graphics accelerator (synonyms: 2D accelerator, 2D accelerator, Windows accelerator or GDI accelerator). Modern window interfaces require a quick (in tenths of a second) redrawing of the screen content when opening/closing windows, moving them, etc., otherwise the user will feel the system’s response to his actions is not fast enough. To do this, the processor would have to process data and transfer it across the bus at a speed only 2-3 times lower than the speed of RAMDAC, and this is tens and even hundreds of megabytes per second, which is practically unrealistic even by modern standards. At one time, to increase system performance, local buses were developed, and later - 2D accelerators, which are specialized graphics processors capable of independently drawing a mouse cursor, window elements and standard geometric shapes on the screen, provided by GDI - the Windows graphics library. 2D accelerators exchange data with video memory over their own bus, without loading the processor system bus. Via the system bus, the 2D accelerator receives only GDI instructions from the central processor, while the volume of transferred data and processor load are hundreds of times less.

Modern 2D accelerators have a 64- or 128-bit data bus, and to effectively use the capabilities of this bus, the video card must have 2 or 4 MB of video memory installed, respectively, otherwise the data will be transferred over a bus that is twice as narrow with a corresponding loss in performance.

We can say that by now 2D accelerators have reached perfection. All of them work so quickly that despite the fact that their performance on special tests may differ from model to model by 10-15%, the user most likely will not notice this difference. Therefore, when choosing a 2D accelerator, you should pay attention to other factors: image quality, availability of additional functions, quality and functionality of drivers, supported frame rates, compatibility with VESA (for fans of DOS games), etc. 2D accelerator chips in currently produced by ATI, Cirrus Logic, Chips&Technologies, Matrox, Number Nine, S3, Trident, Tseng Labs and other companies.

Multimedia accelerators usually mean devices that, in addition to accelerating conventional graphics operations, can also perform a number of operations for processing video data from different sources.

First of all, these are functions to speed up video output in AVI, Indeo, MPEG-1 and other formats. The problem is that video in NTSC format runs at 30 frames per second, PAL and SECAM at 25 frames per second. The frame rate in digital video of the listed formats is also less than or equal to 30 fps, but the image resolution rarely exceeds 320 x 240 pixels. With these parameters, the speed at which information arrives is about 6 MB/s, and the processor manages to decompress it and send it across the bus to video memory. However, this image size is too small for comfortable viewing on the screen, so it is usually scaled to fit the entire screen. In this case, the data flow rate increases to tens and hundreds of megabytes per second. This circumstance has led to the emergence of video accelerators that can independently scale video in AVI and MPEG-1 formats to fill the entire screen, as well as smooth the scaled image so that it does not look like a set of squares. The vast majority of modern 2D accelerators are also video accelerators, and some, such as the ATI Rage128, can also play video in MPEG-2 format (i.e., with an original resolution of 720 x 480).

Multimedia functions also include hardware digital video compression and decompression (which is almost never found on mass-produced video cards), the presence of a composite video output, TV signal output to a monitor, low-frequency video input and high-frequency TV input, a module for working with teletext and other functions.

3D accelerators

When computer games became the engine of progress, 2D accelerators (see Video accelerators) almost exhausted their capabilities, and the evolution of video cards took the path of endowing them with increasingly powerful means of accelerating 3D computer graphics. Video adapters capable of accelerating 3D graphics operations are called 3D accelerators (a synonym is 3D accelerator, as well as the common slang "3Dfx" to refer to all 3D accelerators, not just those manufactured by 3Dfx Interactive). In general, 3D accelerators existed before, but their area of ​​application was three-dimensional modeling and CAD; they were very expensive (from 1 to 15 thousand dollars) and were practically inaccessible to the mass user.

What actions does the 3D accelerator accelerate? In a computer, three-dimensional objects are represented using geometric models consisting of hundreds or thousands of elementary geometric shapes, usually triangles. The spatial position of the light sources, the reflective properties of the surface material of the object, the degree of its transparency, etc. are also specified. In this case, some objects may partially block each other, and light may be reflected between them; the space may not be completely transparent, but shrouded in fog or haze. For greater realism, it is necessary to take into account the effect of perspective. To prevent the surface of the modeled object from looking artificial, a texture is applied to it - a small two-dimensional image that conveys the color and texture of the surface. All of the listed three-dimensional objects, taking into account the effects applied to them, must ultimately be converted into a flat image. This operation, called rendering, is performed by the 3D accelerator.

Let's list the most common operations that a 3D accelerator performs at the hardware level:

Removing invisible surfaces. It is usually performed using the Z-buffer method, which consists in the fact that the projections of all points of a three-dimensional model of an object onto the image plane are sorted in a special memory (Z-buffer) according to the distance from the image plane. The color of the image at a given point is the color of the point in the Z-buffer that is closest to the image plane, and the remaining points are considered invisible (unless the transparency effect is turned on), since they are blocked from us by the very first point. This operation is performed by the vast majority of 3D accelerators. Most modern accelerators have 16-bit Z-buffers located in on-board video memory.

Shading gives the triangles that make up an object a specific color depending on the lighting. It can be uniform (Flat Shading), when each triangle is painted evenly, which causes the effect of not a smooth surface, but a polyhedron; Gouraud Shading, which interpolates color values ​​along each edge, giving curved surfaces a smoother appearance without visible edges; according to Phong Shading, when normal vectors to the surface are interpolated, which makes it possible to achieve maximum realism, but requires large computational costs and is not yet used in mass 3D accelerators. Most 3D accelerators can perform Gouraud shading.

Clipping defines the portion of an object that is visible on the screen and clips the rest to eliminate unnecessary calculations.

Lighting calculation. To perform this procedure, the Ray Tracing method is often used to take into account the reflections of light between objects and their transparency. All 3D accelerators can perform this operation with varying quality.

Texture Mapping, or the overlay of a flat raster image on a three-dimensional object in order to make its surface more realistic. For example, as a result of such an overlay, a wooden surface will look exactly like it was made of wood, and not of an unknown homogeneous material. High-quality textures usually take up a lot of space. To work with them, 3D accelerators on the AGP bus are used, which support texture compression technology. The most advanced cards support multitexturing - simultaneous overlay of two textures.

Filtering and Anti-aliasing. Anti-aliasing refers to the reduction of distortion in texture images by interpolating them, especially at boundaries, while filtering refers to a method of reducing unwanted “grain” when changing the scale of a texture when approaching or moving away from a 3D object. Bilinear filtering is known, in which the color of a pixel is calculated by linear interpolation of the colors of neighboring pixels, as well as better trilinear filtering using MIP maps (Trilinear MIP Mapping). MIP maps (from the Latin Multum in Parvum - “many in one”) are understood as a set of textures with different scales, which allows, in the process of trilinear filtering, to perform averaging between neighboring pixels and between neighboring MIP maps. Trilinear filtering gives a special effect when applying textures to an extended object moving away from the observer. Modern boards support trilinear filtering.

Transparency, or alpha channel of an image (Transparency, Alpha Blending) is information about the transparency of an object, allowing you to construct transparent and translucent objects such as water, glass, fire, fog and haze. Fogging is often separated into a separate function and calculated separately.

Color mixing, or dithering, is used to process two- and three-dimensional images with a large number of colors on a device with fewer of them. This technique involves drawing a special pattern with a small number of colors, which creates the illusion of using a larger number of colors when moving away from it. An example of dithering is a method used in printing to convey gradations of gray color by applying small black dots with different spatial frequencies. 3D accelerators use dithering to render 24-bit color in 8- or 16-bit modes.

To support 3D accelerator functionality in games and other programs, there are several application programming interfaces, or APIs (Application Program Interfaces), that allow an application to use the capabilities of a 3D accelerator in a standard way. Today there are many such interfaces, among which the most famous are Direct3D (Microsoft), OpenGL (Silicon Graphics), Glide (3Dfx), 3DR (Intel), Heidi (Autodesk), RenderGL (Intergraph).

Microsoft's Direct3D interface has become the de facto standard for most computer games; and most 3D accelerators are equipped with Direct3D drivers. However, it is worth keeping in mind that Direct3D is supported only in Windows 95/98, and already in Windows NT, most boards do not support hardware acceleration functions.

Developed by Silicon Graphics for its Iris GL graphics stations, the OpenGL application programming interface has become the generally accepted standard for 3D modeling and CAD programs. Used in professional 3D accelerators, it allows you to very accurately describe the parameters of a scene. OpenGL is currently an open standard controlled by the OpenGL Architecture Review Board, which in addition to Silicon Graphics includes Digital, IBM, Intel, Intergraph, Microsoft, and others. Despite this, there are many dialects of OpenGL. In terms of prevalence in the field of computer games, OpenGL is inferior to Direct3D.

The 3D accelerator driver can support OpenGL in two modes: truncated MCD (Mini Client Driver) and full ICD (Installable Client Driver). The MCD driver implements only a basic set of operations, the ICD is a highly optimized driver that provides maximum performance. Unfortunately, many 3D accelerator manufacturers, while declaring their full OpenGL support, do not provide it even at the MCD driver level. Only a few 3D accelerators (mainly based on 3DPro, Glint, Permedia 2 and RivaTNT chipsets) can boast of having stable ICD drivers.

The Glide interface was developed by 3Dfx Interactive for its Voodoo accelerators. Glide has gained widespread adoption among computer game makers, although, unlike OpenGL, Glide is not a universal 3D API and only supports Voodoo capabilities.

Currently, the most famous 3D accelerators are: ATI 3D Rage Pro and 3D Rage 128; Intel i740; Number Nine Ticket to Ride IV; Mitsubishi 3DPro/2mp, Matrox G100 and G200; S3 Savage3D; Riva128 and RivaTNT; Rendition V2100 and V2200; 3Dlabs Permedia 2 and 3; 3Dfx Voodoo, Voodoo2 and Voodoo Banshee; NEC PowerVR PCX2. On the basis of these chipsets, video cards themselves are produced, not only by the listed companies, but also by companies that do not produce their own graphics processors, for example ASUSTek, Creative Labs or Diamond Multimedia. Modern 3D video cards also have 2D graphics acceleration functions. The exception is accelerators based on 3Dfx Voodoo and Voodoo2, which are connected to the output of a regular video card in front of the monitor with a special external connecting cable. This solution degrades the quality of the 2D image, and it is also impossible to work in windowed mode. In 3Dfx Voodoo Banshee, this scheme was abandoned, and it is a full-fledged 2D/3D accelerator.

Sound card

The sound card was one of the most recent improvements in the personal computer. It is installed in one of the motherboard connectors as a daughter card. And performs computational operations related to the processing of sound, speech, and music. The sound is played through external sound speakers connected to the sound card output. A special connector allows you to send an audio signal to an external amplifier. There is also a microphone connector, which allows you to record speech or music and save it on your hard drive for later processing and use.

The main parameter of the sound card is the bit depth, defining the number of bits used when converting signals from analog to digital form and vice versa. The higher the bit depth, the smaller the error associated with digitization, the higher the sound quality. The minimum requirement today is 16 bits, and 32-bit and 64-bit devices are the most common.

In the field of audio reproduction, the most difficult situation is with standardization. In the absence of uniform centralized standards, devices compatible with the device have become the de facto standard SoundBlaster trademark owned by the company Creative Labs.

Recently, audio processing has been viewed as a relatively simple operation, which, due to the increased power of the processor, can be entrusted to it. In the absence of increased requirements for sound quality, you can use integrated sound systems, in which audio processing functions are performed by the central processor and motherboard chips. In this case, speakers or other audio playback devices are connected to sockets installed directly on the motherboard.

Systems located on the motherboard

Random Access Memory (RAM *-- Random Access Memory) -- it is an array of crystalline cells capable of storing data. There are many different types of RAM, but from the point of view of the physical principle of operation they distinguish dynamic memory (DRAM) And static memory (SRAM).

Dynamic memory (DRAM) cells can be represented in the form of microcapacitors capable of accumulating charge on their plates. This is the most common and economically available type of memory. The disadvantages of this type are associated, firstly, with the fact that both when charging and discharging capacitors, transient processes are inevitable, that is, data recording occurs relatively slowly. The second important drawback is related to the fact that cell charges tend to dissipate in space, and very quickly. If the RAM is not constantly “recharged,” data loss occurs within a few hundredths of a second. To combat this phenomenon, the computer constantly regeneration (refreshing, recharging) RAM cells. Regeneration occurs several tens of times per second and causes wasteful consumption of computing system resources.

Static memory cells (SRAM) can be thought of as electronic trace elements -- triggers, consisting of several transistors. A trigger stores state, not charge. (on/off), therefore, this type of memory provides higher performance, although it is technologically more complex and, accordingly, more expensive.

Dynamic memory chips are used as the main RAM of a computer. Static memory chips are used as auxiliary memory (the so-called cache memory), designed to optimize processor performance.

Each memory cell has its own address, which is expressed as a number. In most modern processors, the address size limit is usually 32 bits, which means that there can be a total of 2 32 independent addresses. One addressable cell contains eight binary cells in which 8 bits, that is, one byte of data, can be stored.

Thus, in modern computers it is possible direct addressing to a memory field of size 2 32 bytes = 4 GB. However, this does not mean that this is exactly how much RAM a computer must have. The maximum size of the RAM field installed in the computer is determined by the microprocessor kit (chipset) motherboard and usually cannot exceed several GB. The minimum amount of memory is determined by the requirements of the operating system and for modern computers is 128 MB.

An idea of ​​how much RAM there is there must be in a typical computer, changes continuously. In the mid-80s, a memory field of 1 MB seemed huge; in the early 90s, 4 MB was considered sufficient; by the mid-90s it increased to 8 MB, and then to 16 MB. Today, the typical RAM size is 256 MB, but the upward trend continues.

RAM in a computer is located on standard panels called modules. RAM modules are inserted into the corresponding slots on the motherboard. If you have easy access to the connectors, you can do the operation yourself. If there is no convenient access, partial disassembly of the system unit components may be required, and in such cases the operation is entrusted to specialists.

The main characteristics of RAM modules are memory capacity and data transfer speed. Today, the most common modules are 128-512 MB. The data transfer rate determines the maximum memory bandwidth (in MB/s or GB/s) in the optimal access mode. This takes into account memory access time, bus width and additional capabilities, such as transmitting multiple signals in one clock cycle. Modules of the same volume can have different speeds...........