Understanding about the calculation technique. Personal computers Functional scheme of the computer. Warehouse for accounting equipment of complexes in the arm
Understanding the Counting System
A specific set of attachments and programs that interact with each other, appointments for servicing one working business, is called counting system.
The collection of outbuildings, which are used for automatic or automated processing of data, is called counting technique or hardware problems. The warehouse of the counting system is called configuration. The hardware and software of the computational system is taken to be seen in detail.
Criteria for choosing a hardware and software solution productivityі efficiency.
The central attachment of the computational system is a computer. Computer- a whole electronic attachment, intended for automation of the creation, storage, processing and transportation of data. If you want to look at the hardware and software features in a clear way, you should indicate that these features of the calculation system work in an uninterrupted connection and in an uninterrupted relationship.
History of the development of numerical technology. Generation of EOM.
(Proponuetsya to students for self-studying).
EOM classification
The mass character of the PC is not guilty of standing up for the fact that the Crimean PC is and other more difficult calculation systems:
· Supercomputers;
Large EOM (mainframes);
· Minicomputer;
· Microcomputers (personal PCs are brought before them).
Qi EOM vіdrіznyayutsya:
· Productivity;
Rosemary;
functional recognition.
Architecture and structure of EOM
When looking at the computer outbuildings, it was decided to distinguish between their architecture and structure.
architecture The computer is called yogo description in the singal common language, which includes a description of the possibilities of programming, command systems, addressing systems, memory organization, etc. The architecture defines the principles of dividing, informational links and mutually connecting the main logical nodes of the computer: processor, operational memory, external memory and peripheral devices. Spіlnіst arkhіtektury other computers I’ll take care of my sense of confusion at the glance of a koristuvach.
Structure A computer is a collection of its functional elements and links between them. Elements can be of different extensions - from the main logical knots of the computer to the simplest schemes. The structure of the computer is graphically represented as block diagrams, with the help of which it is possible to describe the computer in any detail.
Classic architecture(von Neumann architecture) - one arithmetic-logical extension (ALU), through which the flow of data passes, that one control extension (CU), through which the flow of commands passes - the program. Tse single processor computer. Which type of architecture is considered to be the architecture of a personal computer hot tire. All functional blocks here are tied together with a hard tire, also called system highway. Physically, the main line is a rich-wire line with sockets for connecting electronic circuits. The number of wires in the backbone is subdivided into groups: address bus, data bus and control bus.
Classification of methods of calculation technique
1. Hardware security
The warehouse of the counting system is called a configuration. The hardware and software for the computational technique is taken to be seen in detail. Apparently, it is clear to look at the hardware configuration of the enumeration systems and the software configuration. Such a principle can be especially important for informatics, but the shards are often more successful than the workers themselves can be taken care of both by hardware and software tools. The criteria for choosing a hardware and software solution are productivity and efficiency. Please note that hardware solutions are more expensive on average, but the implementation of software solutions will require more highly qualified personnel.
Before hardware security of counting systems, there are attachments and accessories that fix the hardware configuration. Modern computers and calculation complexes may have a block-modular design - hardware configuration, necessary for the development of specific types of robots, as it is possible to select from ready-made nodes and blocks.
The main hardware components of the computational system are: central processor that peripheral attachments, which are connected between themselves by the system backbone (Fig.1.) As a rule, the difference between the middle becomes 1 byte. Typical operations on the main memory: reading that record instead of the middle address.
2. CPU
The central processor is the central attachment of the computer, which is responsible for the operation of data processing and processing by the peripheral devices of the computer. To the warehouse of the central processor enter:
Attachment management - organize the process of vikonannya program and coordinate the interaction of all attachments in the accounting system for hours and hours of work;
Arithmetic-logical attachment - vikonu arithmetic that logical operations over tributes: addition, vіdnіmannya, multiple, rozpodіl, porіvnyannya that іn;
Memorable attachment - є internal memory the processor, which is composed of registers, in case of any change, the processor wins the calculation and saves the intermediate results; to speed up the work with the operational memory, the cache-memory is broken;
Clock frequency generator - generates electrical impulses that synchronize the operation of all nodes of the computer.
The central processor monitors various operations with data for additional specializations for the collection of key changes and hourly results - internal registers. Registers are divided into two types (Fig. 2.):
Registers of global recognition - are selected for time-based collection of key local changes and intermediate results, including data registers and indicator registers; the main function of the safety swedish access to often victorious tributes (sound without ringing to memory).
Specialized registers - win the control of the processor's robots, the most important of them: the command register, the stack indicator, the flag register and the register, to retrieve information about the program camp.
The register of these programmers can win on your own account for the time-based collection of any objects (given or the address) and perform the necessary operations on them. Index register just like the register of data, can win with a sufficient rank; їhnє osnovne prichennya - zberіgati іndexi or zmіschennya danih і commands vіd vіd base ї address (when choosing operands from memory). Base addresses in your own can be in base registers.
Segment registers are the most important element of the processor architecture, ensuring that the 20-bit address space is addressed behind the additional 16-bit operands. Main segment registers: CS – segment code register; DS – data segment register; SS – stack segment register, ES – additional segmental register. Zvernennya to memory zdіysnyuєtsya at vyglyadі segmentіv - logical utvornіh, overlays be-yakі dіlyanki physical address space. Pochatkov address of the segment, subdivided by 16 (without the youngest sixteenth digit) shall be entered in one of the segment registers; after which access to the memory box is expected, which starts from the specified segment address.
The addresses, whether in the middle of the memory, are composed of two words, one of them is marked in the memory of the second segment, and otherwise - used in the boundaries of the second segment. The size of the segment is determined by the agreement of data, which is located in the new, but in no case can you exceed the value of 64 Kbytes, which is determined by the maximum possible size of the sound. Segment addresses of the command segment are stored in the CS register, and zsuv to the byte to be addressed - in the IP command indicator register.
Fig.2. 32-bit processor registers
After the program is entered into the IP, the first command of the program is entered. The processor, having taken into account the memory, changes the IP exactly by the length of the command (the commands of the Intel processors can take the maximum length of 1 to 6 bytes), the result of the IP is shown on the other program command. After typing the first command, the processor reads from another memory, and renews the greater IP value. As a result, the IP will always contain the replacement of the black team - the team that follows the win. The descriptions of the algorithm are only broken for the typing of transition commands, calling subprograms and service interruptions.
The segment addresses of the data segment are stored in the DS register, but they can be in one of the registers of the global recognition. The additional segment register ES is selected for data entry before data watering, which does not enter programs, for example, a video buffer or system cores. However, due to the need, yoga can be adjusted for one of the segments of the program. For example, as the program works with a great deal of data, you can transfer two segments for them and enter up to one of them through the DS register, and up to the last one through the ES register.
The SP stack indicator register is counted as the stack top indicator. The stack is called the area of the program for the timely saving of certain data. The stability of the stack is due to the fact that this area wins out differently, moreover, the savings in the stack of data and the selection of data are victorious for the help of push and pop commands without specifying names. The stack is traditionally victorious for saving the registers that are victorious by the program, before the weekly subprogram, so, at your own hand, victorious register of the processor for your own special purposes. The last time the registers appear from the stack after turning from the subprogram. The second extension of the reception is the transfer of subprograms of the parameters it needs through the stack. The subprogram, knowing in what order the parameters are placed on the stack, can pick up their stars and beat them for the hour.
The peculiarity of the stack is its own order of selection of data, which should be left in the new: at any moment in the stack only the top element is available, that element, there will be no more entanglements in the stack. Vivantage from the stack of the upper element to defeat the accessible offensive element. Elements of the stack are sorted in the memory area, entered under the stack, starting from the bottom of the stack (from the first maximum address) behind the addresses that change sequentially. The addresses of the top accessible element are stored in the SP stack indicator register.
Special registers are only available to privileged users and are managed by the operating system. They control various blocks of cache memory, main memory, add-on input and output and other add-ons of the computational system.
Use one register, which is available both to a privileged person, and in the mode of koristuvach. Ce register PSW (Program State Word - the word will become a program), which is called the great-principal. The flagship register to avenge different bits, which are necessary for the central processor, the most important - codes of minds, which are victorious at the same time of mental transitions. The stench is installed in the skin cycle of the arithmetic-logic processor and reflects the result of the previous operation. In the place of the ensign register to deposit according to the type of calculation system and it can include additional fields, which indicate: the mode of the machine (for example, koristuvalnitsky or privileges); bit trasuvannya (which vikoristovuєtsya for nalagodzhennya); equal to processor priority; the status is allowed to be interrupted. The flag register is read in the koristuvach mode, but other fields can only be written in the privileged mode (for example, bit, which mode you specify).
Command indicator register After the choice of the command, remember the register of commands to be corrected and the indicator to go to the next command. The indicator of commands follows the progress of the program, showing at the moment the address of the command that is following the command. Registry of software inaccessible; increasing the address of the new vicon microprocessor, vrakhovuychi at his dozhina flow command. Commands for switching, restarting, calling the subprogram and turning them are changed instead of the indicator, then go to the necessary points of the program.
The battery register wins in the most important number of commands. Teams that often zastosovuyutsya, vikoristovuyut tsey register, shortening the format.
To process information, organize the transfer of data from the middle memory to the register of global recognition, the operation by the central processor and the transfer of results to the main memory. Programs are saved as a sequence of machine commands, so that they can override the central processor. The skin command is formed from the field of operation and the field of operands - data, over which ones are vikonuetsya given operation. The set of machine commands is called machine language. Vikonannya program zdіysnyuєtsya in this way. The machine command, as shown by the program id, is read from memory and copied to the command register, decoded, after which it is written. After the next program stop, it indicates the coming command, etc. Tsі dії are called a machine cycle.
More central processors have two modes of operation: the kernel mode is the core mode, which is set by the bit of the word I will become a processor (ensign register). If the processor is running in kernel mode, it can override all commands in a set of inputs and override all hardware capabilities. The operating system operates in kernel mode and provides access to all control. Programs of koristuvachіv pratsyyut in the regimen of koristuvach, which allows the typing of anonymous commands, but only a part of the hardware is accessible.
To communicate with the operating system, the program is responsible for generating a system call, which secures the transition to the kernel mode and activates the functions of the operating system. The trap command (emulation interruption) switches the robotic processor mode from the core mode and transfers control to the operating system. After the completion of the robotic control, it turns to the program of the coristuvacha, to the command to follow the system prompt.
On computers, there are instructions for victoring system calls, which are reordered, which are called hardware for warning about faulty situations, for example, try to go to zero or rewind during operations with a floating point. In all similar ways, the control goes to the operating system, as it is the fault of the virishity that the work was done. Sometimes you need to complete the program with notifications about the pardon, sometimes you can ignore it (for example, if you add the value of the number of yogo, you can take it equal to zero) or pass the keruvannya to the program to process some kinds of minds.
According to the method of roztashuvannya pristroїv schodo central processor razreznyayut vnutr_shnі and zvnіshnі podstroї. Zovnіshnіmi, as a rule, are more outbuildings for the introduction of data (they are also called peripheral outbuildings) and deyakі outbuildings, recognized for the trivial collection of data.
Uzgodzhennya mizh okremimi nodes and blocks vykonuyuut behind the help of transitional hardware-logical outbuildings, as they are called hardware interfaces. Standards for hardware interfaces in computing technology are called protocols - a collection of technical minds, which can be provided by retailers of outbuildings for the successful use of their work with other outbuildings.
Numerical interfaces, present in the architecture of any numerical system, can be mentally divided into great groups: succession and parallel. Through the last interface, data is transmitted sequentially, bit by bit, and through the parallel one - at once by groups of battles. The number of battles, which have one message, is determined by the size of the interface, for example, eight-bit parallel interfaces transmit one byte (8 bits) one cycle.
Parallel interfaces can sound foldable attachments, lower last ones, but provide more high productivity. Stop there, it is important to transfer data: for connection of other outbuildings, building outbuildings graphic information, add data recording to your phone, etc. The productivity of parallel interfaces is measured in bytes per second (bytes/s; Kbytes/s; Mbytes/s).
Powering the last interfaces is simpler; as a rule, they do not need to synchronize the robot of the transmitting and receiving attachments (they are often called asynchronous interfaces), but the throughput of their building is less and the coefficient of core power is lower. Oscilki data exchange through the last extensions is carried out not by bytes, but by bits, their productivity is reduced by bits per second (bit / s, Kb / s, Mb / s). Regardless of the simplicity, I will shift the single shifting speed of the sequential transmission into the single shifting speed of the parallel transmission of data with a mechanical path to 8, such a shift cannot be beaten, the shards of faults are not correct through the presence of service data. In extreme cases, with an amendment for service data, sometimes the speed of the last additions is in signs for a second or symbols for a second (s / s), but the value may not be technical, but a dovіdkovy, calm character.
Послідовні інтерфейси застосовують для підключення повільних пристроїв (найпростіших пристроїв друку низької якості: пристроїв введення та виведення знакової та сигнальної інформації, контрольних датчиків, малопродуктивних пристроїв зв'язку тощо), а також у тих випадках, коли немає істотних обмежень щодо тривалості обміну даними (Цифрові cameras).
The other main storage computer is memory. The memory system is designed as a hierarchy of balls (Fig. 3.). The top ball is formed from the internal registers of the central processor. Internal registers allow for saving 32 x 32 bits on a 32-bit processor and 64 x 64 bits on a 64-bit processor, which is less than one kilobyte in both modes. The programs themselves can manage the registers (to check, what to save from them) without the need to insert equipment.
Fig.3. Typical hierarchical memory structure
The offensive ball has a cache memory, it is more importantly controlled by the owner. RAM is divided into cache rows, 64 bytes each, addressing from 0 to 63 for the zero row, from 64 to 127 for the first row, etc. Most of the cache rows are saved from the high cache memory, the middle of the central processor, or even close to it. If the program is guilty of reading the word from memory, the cache chip will check what row the cache needs. As such, it is effectively loaded to the cache memory, the request is satisfied by the cache and the request to the memory is not installed on the bus. In the distance, the ringing to the cache, ring out, at the hour it takes about two beats, and not far away, to lead to the ringing to the memory with a significant second hour. The cache-memory is rimmed at the rozmіrі, which is zimovleno її vysokoy varіstyu. Some cars have two or three equal caches, moreover, the skin is more advanced and larger behind the front.
Dalі sleduє operative memory (RAM - operative attachment, scho zam'yatovuє, English. RAM, Random Access Memory - memory with enough access). The main working area of the memory of the enumeration system. Usі zapiti central processor, yakі mozhut buti vykonnі kesh-memory, nadhodit for processing in the main memory. When using a number of programs on the computer, it is necessary folding programs place in operational memory. Protecting programs from each other that yogo relocation in memory is implemented with the help of setting up a computer with two special registers: the base register and the boundary register.
In the simplest way (Fig. 4.a), if the program is starting to work, the addresses of the top of the program module are taken into the base register, which is to be fixed, and the boundary register is to talk about those, which are borrowed from the start of the program module at once with data. When you select a command from memory, the equipment recalculates the name of the commands, as well as the smaller, lower boundary register, adding to the new value of the base register, and transferring the amount to memory. If the program wants to read the word danikh (for example, from the address 10000), the equipment automatically adds the address to the base register (for example, 50000) and transfers the sum (60000) to memory. The base register gives the program the ability to ask for a part of the memory, what to go for the address, what to save in the new. In addition, the border register keeps the programs from being executed until the memory of the programs is still part. In such a rank, behind the help of the scheme, offenses are being lied to: zahistu that relocation of programs.
As a result of the re-verification, that re-writing of data, addresses, formed by the program is called a virtual address, translated to the address, as the memory is victorious, and it is called a physical address. Attachment, which vikonu reverb that transformation, is called an attachment for memory care or a memory manager (MMU, Memory Management Unit). The memory manager is known either by the processor circuit, or close to it, but it is logical to be between the processor and the memory.
A larger foldable memory manager is folded over two pairs of base and boundary registers. One pair is assigned to the text of the program, another pair is assigned to the data. The command register and sending to the text of the program are processed from the first pair of registers, sending a pair of registers to a friend. The reason for such a mechanism is the possibility of expanding one program between dekilkom koristuvachami when saving in RAM only one copy of the program, as in a simple scheme. Working program No. 1 chotiri register is roztashovuyutsya so, as shown in Fig. 4 (b) Zlіva, pіd hour of work program No. 2 - on the right. Handled by the memory manager is a function of the operating system.
Let's step at the memory structure magnetic disk(hard disk). Disk memory is two orders of magnitude cheaper for RAM and a bit more expensive for size, while access to data hosted on disk takes about three orders of magnitude more than an hour. The reason for the low speed hard drive The fact that the disk is a mechanical design. hard drive it is made up of one or more metal plates, which are wrapped around 5400, 7200 or 10800 revolutions per piece (Fig. 5.). The information is recorded on the plate at the concentric keels. The read/write heads in the skin position can read the ring on the plate, which is called a track. All the tracks at the same time for a given position of the fork fix the cylinder.
The skin track is divided into a few sectors, sound 512 bytes per sector. On modern disks, clean cylinders Great number sectors, lower internal. The movement of the head from one cylinder to the next loan is close to 1 ms, and the movement to the second cylinder takes 5 to 10 ms to fall on the disc. If the head is rotting over the required track, follow the check, while the engine turns the disk so that it becomes the necessary sector under the head. Tse borrow additionally from 5 to 10 ms fallowly from the speed of the disc wrap. If the sector is reloaded under the head, the process of reading or writing is taken from 5 MB / s (for low-speed disks) to 160 MB / s (for high-speed disks).
The remaining ball is occupied by a magnetic line. Tsey wearer often victorious for the meeting backup copies expanse of a hard disk or for saving great data sets. For access to the information, the line was placed on an attachment for reading magnetic lines, then it was rewound to the input block with the information. The whole process is hvilini. A typical memory hierarchy is described, but in some variants there may be not all equals, but others can be seen (for example, an optical disk). In the case of the Russian ієrarchії beast down the hour of sufficient access, it will significantly increase in the annex to the annex, and the place of growth is equivalent to the hour of access.
Krіm descriptions are more visible on rich computers є permanent memory with sufficient access (ROM - permanent attachment, ROM, Read Only Memory - memory only for reading), as it does not waste its time when the system is alive. The ROM is programmed in the process of modification, and the latter cannot be changed in the same place. On some computers in the ROM, there are programs for the cob capture, which are activated when the computer is started, and there are I/O cards for keruvannya with low attachments.
Electrically erased PZP (EEPROM, Electrically Erasable ROM) and flash RAM (flash RAM) are also energy-independent; However, it will take more than an hour to write data to them, less to write to the RAM. That's why the stench vikoristovuyutsya so very, like ROM.
There is one more type of memory - CMOS memory, which is energy-saving and vicorous for saving streaming dates that current hour. Memory otrimu zhivlennya vіd akkumulyatornogo vbudovanogo in kom'yuter, can mіstiti parameters configuration (for example, vkazіvka, from any hard drive robita zavantazhennia).
3. Add-on
The other annexes, which closely interact with the operating system, are the addition of an input-output, as they are composed of two parts: I will add the controller itself. The controller is a microcircuit (a set of microcircuits) on the board, which is inserted into the sockets, so that it receives and viconuates the commands of the operating system.
For example, the controller receives a command to read the first sector of the disk. In order to complete the command, the controller converts the linear number of the disk sector to the number of the cylinder, sector, and head. The transformation operation is made more complicated by the fact that the outer cylinders can have more sectors, lower internal ones. Then the controller selects a head over which cylinder at a given moment, and gives a sequence of impulses to move the head to the required number of cylinders. After that, the controller checks until the disk turns, having placed the required sector under the head. Then, sequentially, the processes of reading and saving battles in the world from the disk, the process of removing the header and calculating the control sumi. Then the controller chooses to take away bits from the word and takes them from memory. For the purpose of robotic controllers, replace them with microprograms.
The I/O device itself has a simple interface, which can comply with the single IDE standard (IDE, Integrated Drive Electronics - an interface for storage devices). If the interface is attached by the controller, then the operating system will only need to process the interface of the controller, which can be added to the interface.
So how are the controllers for other outbuildings I/O is driven by one type of one, then to manage them you need a separate software security - drivers. The reason for this is that the controllers are responsible for supplying drivers for supporting them with operating systems. To install the driver into the operating system, there are three ways:
Rebuild the kernel with a new driver and then reconfigure the system, so pratsyuє impersonal UNIX systems;
Create a record in the file that enters the operating system, about those that need a driver and re-install the system, know about the hour of the start-up of the operating system required driver ta zavantazhit yogo; this is how the Windows operating system works;
Adopt new drivers and promptly install them with the help of the operating system for an hour of work; Able to overcome USB and IEEE 1394 busses, which will always require dynamically driven drivers.
For communication with the skin controller, register number. For example, a minimal disk controller can register to assign a disk address, a memory address, a sector number, and a direct operation (read or write). To activate the controller, the driver removes the command from the operating system, then translates it into the value that is attached to the registry entry.
On some computers, the registers of input-visualization are displayed in the address space of the operating system, so you can read them or write them down as the first words in memory. The addresses of the registers are located in the RAM beyond the limits of the reach of the programs of the coristuvachs, so that the programs of the coristuvachs secure access to the equipment (for example, for the help of the base and boundary registers).
On other computers, the registers of attachments are placed in special input-output ports, and the register may have its own port address. On such machines, the privileged mode has IN and OUT commands available to allow drivers to read and write registers. The first scheme uses the need for special commands for the introduction-vivodu, and also for the vikoristovu deyaku kіlkіst address space. Another scheme does not stick to the address space, but it takes into account the presence of special commands. Offensive schemes are widely vindicated. Introducing and seeing these data is done in three ways.
1.Korystuvatska program is a type of system power, which kernel translates from a call to the procedure of a different driver. Let the driver start the process of entering and exiting. At the same time, the driver has a very short program cycle, I’ll gradually test the readiness to build it, with some wines to practice (sound like a bit, which indicates those who are interested in employment). After the operation is completed, the driver enters the data where necessary, and turns to holiday camp. Then the operating system will turn the program around, as the wake-up call was made. This method is called readiness checks or active checks and may have one shortfall: the processor is guilty of picking up attachments, the wine docks do not complete their work.
2. The driver starts up and asks you to see a reboot after the completion of the introduction. Once the driver turns the data, the operating system will block the call program, as necessary, and begin to perform other tasks. If the controller detects the transfer is completed, it generates a retransmission to signal the completion of the operation. The mechanism for the implementation of the introduction-withdrawal is considered to be such a rank (Fig. 6.a):
Krok 1: the driver sends a command to the controller, writing information to the register; launch controller
Krok 2: after the end of reading or writing, the controller sends a signal to the microcircuit of the controller to reset.
Krok Z: if the controller is ready to receive a reset, it sends a signal to the first contact of the central processor.
Krok 4: the controller resets the number and attaches the entry-view to the bus so that the central processor can read it and recognize it, which attachment has completed the robot. When the central processor accepts a change in the command list (PC) and the processor word (PSW) is placed on the stream stack, and the processor is switched to the privileged robot mode (the kernel robot mode of the operating system). I’ll add a number to the input-vivodu, which can be sorted out as an index of a part of the memory, which can be sorted out to search for the address of the collector, I’ll sort out this add-on. Tsya part of the memory is called the vector of remembrance. If I rebuild the processor (I’ll build a part of the driver, I’ll rebuild it) repair my robot, I’ll see the list of commands in the stack and the word will become the processor, pick up the attachments, so that I can take information about the yoga stan. As soon as the processing is completed, the control turns to the program of the coristuvacha, to the next command, which has not yet been completed (Fig. 6 b).
3. For the introduction-viewing of information, there is a direct memory access controller (DMA, Direct Memory Access), which manages the flow of battles between the operational memory and other controllers without the constant insertion of the central processor. The processor calls the DMA chip, it seems, how many bytes it needs to transfer, it tells you the addresses to add that memory, as well as direct data transfer and allows the chip to work on its own. After completion Robotic DMA initsiyuє pereprivannya, as if they are being turned into a senior rank.
Pererivannya can be taken at inopportune moments, for example, at the hour of the completion of another reparation. Because of this, the central processor may be able to hinder the reboot and allow it later. Until the fence is interrupted, all the extensions, if they have completed the robot, continue to overpower their signals, but the processor robot is not interrupted by the dot, docks will not be allowed to be interrupted. How to finish the robot in the time of the sprinkling of attachments at that hour, if the fence is interrupted, the controller is interrupted, since they can be cut off first, sound spiraling on static priorities, designated for skin attachment.
The Pentium system has a number of buses (cache bus, local bus, memory bus, PCI, SCSI, USB, IDE and ISA). A leather tire has its own transmission speed and its own functions. In the operating system for keruvannya computer that yogo configuration of faults are known about all tires.
ISA bus (Industry Standard Architecture, industry standard architecture) - first appeared on IBM PC / AT computers, working at a frequency of 8.33 MHz and can transmit two bytes per cycle with a maximum speed of 16.67 MB / s .; won is included in the system of zvorotnoї sumіsnostі zі old povіlnimi boards of input-output.
PCI bus (Peripheral Component Interconnect, Peripheral Component Interconnect) - created by Intel as a successor to the ISA bus, can operate at a frequency of 66 MHz and transfer 8 bytes per cycle with a speed of 528 MB / s. In Denmark, the PCI bus has more high-end I/O attachments, as well as computers with processors running Intel, so there are a lot of I/O boards with it.
The local bus in the Pentium system is used to transfer data from the PCI bridge chips to the central processor, which is reduced to the memory of the seen memory bus, which often works at a frequency of 100 MHz.
The cache bus is used to connect the outer cache, the chips of the Pentium system can use the cache of the first level (cache L1), the processor, and the great cache of the other level (cache L2).
The IDE bus is used for attaching peripheral devices: disks and devices for reading CDs. The bus is the interface of the disk controller on PC / AT, in this hour it is included in the standard set of all systems based on Pentium processors.
The USB bus (Universal Serial Bus, universal serial bus) is recognized for the connection to the computer of the main devices for input and output (keyboards, mice). There is a victorious little chotiriprovidny rose, two darts of which supply electricity to the USB devices.
The USB bus is a centralized bus, for which the main attachment is to the skin for a millisecond, it is tested for input and output, for recognition, and data. Won mozhe cheruvati zavantazhennym danikh іz shvidkіstyu 1.5 MB / s. All USB devices use the same driver, so they can be brought to the system without re-engineering.
The SCSI bus (Small Computer System Interface, system interface of small computers) is a highly productive bus that is used for hard drives, scanners and other devices that require a significant bandwidth. Її Productivity is 160 MB/s. The SCSI bus is popular on Macintosh systems, popular on UNIX systems and other systems based on Intel processors.
The IEEE 1394 (FireWire) bus is a bit-serial bus and supports packet data transmission with a speed of up to 50 MB/s. This authority allows the connection of portable digital video cameras and other multimedia attachments to a computer. On the USB bus view, the IEEE 1394 bus does not have a central controller.
The operating system is responsible for recognizing the hardware warehouses and for replacing them. This helped the companies Intel and Microsoft to develop a personal computer system called plug and play (plug and play). Prior to the advent of the system, the skin board for input-output is small fixed addresses of input-output registers and the cost of rewriting. For example, the keyboard victorious override 1 that addresses in the range from 0x60 to 0x64; floppy disk controller vikoristovuvav 6 those addresses from 0x3F0 to 0x3F7; the printer was resetting 7 and with addresses from 0x378 to 0x37A.
Yakshcho koristuvach kupuvav sound card and the modem, traplyalos, scho tsі pristroї vipadkovo vykoristovuvali one tezh pereprivannya. Vinikav conflict, it was impossible to build it at once. Possible solutions it was necessary to set the DIP jumpers (jumpers, jumper) into the skin board and adjust the skin board in such a way that the port addresses and numbers were changed other outbuildings did not conflict one with one.
Plug and play allows the operating system to automatically pick up information about input-output devices, centrally recognize the address of the input-output address, and then update the skin board for this information. Such a system is used on Pentium computers. Leather computer with a Pentium processor to replace the motherboard, on which the program is known BIOS system(Basic Input Output System - basic input-output system). BIOS to retaliate with low-level entry-view programs, including procedures: for reading from the keyboard, for displaying information on the screen, for entering-viewing data from the disk and in.
Under the hour of the start of the computer's advance, the BIOS system starts up, as it checks the number of installed in the system of operational memory, the connection and the correctness of the robotic keyboard and other main attachments. Then the BIOS checks the ISA and PCI buses and all attachments attached to them. Acts of tsikh pristroїv traditsіynі (created before the exit to the plug and play standard). They may fix the address of the input-output port (for example, tasks for additional jumpers or jumpers on the input-output board without the possibility of changing them by the operating system). Qi addons are registered, let the registration go through plug and play. If the existing outbuildings are winding down in silence, when there was an hour for the rest of the excitement, then the configuration of new outbuildings.
Let's choose the BIOS settings, for which you will be interested, according to the skin of the list, which is stored in the CMOS memory. Koristuvach can change the list, upgrade to the configuration BIOS program immediately after the vanity. Ring on the back of your head to try to get occupied with the floppy disk. Just don't go into it, win the CD. Whether the computer has a hard disk or a CD, the system is driven from the hard disk. Z zavantazhuvalnogo pristroyetsya in the memory that vykonuetsya first sector. In this sector there is a program that rechecks the table of distributions for example of a promising sector, to determine which of the distributions is active. Let's read the second zavantazhuvach later. Vіn zchituє z active partition of the operating system and start it up.
Once the operating system is supported by the BIOS, it takes information about the configuration of the computer and changes the presence of the driver for the skin device. If the driver is disabled, the operating system will ask the driver to insert a floppy disk or a CD with the driver (i disks will be supplied by the add-on picker). Just like all the drivers at home, the operating system is wrapped up in the kernel. Then we will initialize the driver tables, create all the necessary background processes and run the password entry program. graphical interface skin terminal.
5. The history of the development of methods for calculating the technique
All IBM-sum personal computers are equipped with Intel-sum processors. Briefly, the history of the development of microprocessors of the Intel family is as follows. The first universal microprocessor from Intel appeared in 1970. The wine was called Intel 4004, which used to be chaotic and capable of input/output and the processing of chotiribite words. Svidkodiya yoga became 8000 operations per second. The microprocessor Intel 4004 is designed to be used in programming calculators with a memory size of 4 Kbytes.
Three years later, Intel released the 8080 processor, which can instantly handle 16-bit arithmetic operations, can also address a 1b-bit address bus, and can also address up to 64 KB of memory (25160 = 65536). 1978 was marked by the release of the 8086 processor with a word size of 16 bits (two bytes), a 20-bit bus and an instantaneous operation with 1 MB of memory (2 520 0 = 1048576, or 1024 KB), divided into blocks (segments) of skins. The 8086 processor was equipped with computers that combined with the IBM PC and the IBM PC / XT. We will come to the great crack in the development of new microprocessors, becoming the 8028b processor, which appeared in 1982. Vіn mav 24-bit address bus, instantaneous distribution of 16 megabytes of address space and put on a computer, combined with an IBM PC / AT. In the early 1985 year, the 80386DX was released with a 32-bit address bus (the maximum address space is 4 GB), and in the red 1988 year, the 80386SX was released, the more cheap version was 80386DX and 24-bit. Then in April 1989, the 80486DX microprocessor was announced, and in January 1993, the first version of the Pentium processor (offended by the 32-bit address bus).
In early 1995, at the international exhibition Comtek-95 near Moscow, Intel presented a new processor - P6.
One of the most important goals set by the P6 expansion was a boost in productivity against the Pentium processor. At any rate, the first versions of P6 will be backed up by the already upgraded "Intel" remaining versions Pentium conductor technology (0.6 µm, Z, Z).
Using the same brewing process gives you a guarantee that the mass brewing of the P6 will be fine-tuned without serious problems. At the same time, this means that the sub-productivity is only accessible for the account of the overall improvement of the microarchitecture of the processor. During the development of microarchitecture P6, a combination of various architectural methods was carefully thought out and perfected. Some of them were previously tested in the processors of "great" computers, some were propagated by academic institutions, developed by engineers from the Intel company. This unique combination of architectural features, which in "Intel" is called "dynamic vikonannya", allowed the first P6 crystals to topple the level of productivity, which was planned on the back.
When paired with alternative "Intel" processors of the x86 family, it is clear that the microarchitecture of the P6 may be richly connected with the microarchitecture of the processors Nx586 of the NexGen company and the K5 of the AMD company, and the smaller world, with the "Cyrix" company. This controversy is explained by the fact that the engineers of several companies were violating the same task: the selection of elements of RISC technology while conserving intelligence with the CISC architecture of Intel x86.
Two crystals in one case
The main advantage and the unique feature of P6 are placed in one case with the processor, the secondary static cache memory is 256 kb in size, connected from the processor by a specially seen bus. Such a design can significantly simplify the design of systems with the R6 improvements. P6 is the first purpose for mass production of a microprocessor, which can accommodate two chips in one package.
Krystal CPU R6 to achieve 5.5 million transistors; crystal cache memory of another equal - 15.5 million. For comparison, the remaining Pentium model included about 3.3 million transistors, and the cache memory of another equal was implemented behind an additional set of memory crystals.
Such a large number of transistors in cache is explained by its static nature. The P6 has a static memory of six transistors to store one bit, while the dynamic memory would have enough one transistor per bit. The memory of shvidsha is static, but dearer. If the number of transistors on a chip with a secondary cache is larger, lower on a processor chip, the physical size of the cache is smaller: 202 square millimeters versus 306 for the processor. Offset crystals are placed in a ceramic case with 387 contacts (dual cavity pin-drid array). Obidva crystals are viroblyayutsya іz zastovannyam odnієї i ієї zh tekhnologii (0.6 μm, 4-ball metal-BiCMOS, 2.9 V). Possible maximum energy saving: 20 W at 133 MHz.
The first reason for combining the processor and the secondary cache in one package is the ease of design and flexibility of high-performance systems based on the P6. The productivity of the computational system, based on a Swedish processor, is already due to the exact adjustment of microcircuits for sharpening the processor, the back-end cache. Not all computer maker companies can afford to follow up. In P6, the secondary cache has already been upgraded to the optimal processor rank, which makes it easier to design the motherboard.
Another reason for the consolidation is the increase in productivity. Kzsh of another level of overlap with the processor is specially seen with a bus width of 64 bits and works at the same clock frequency as the processor.
The first Pentium processors with a clock frequency of 60 and 66 MHz were converted to the secondary cache on a 64-bit bus with a clock frequency. It has become more and more difficult and expensive for designers to increase the clock frequency of Pentium on the motherboard. To that began to zastosovuvat dіlniki frequency. For example, in a 100 MHz Pentium, the bus runs at a frequency of 66 MHz (in a 90 MHz Pentium, it is usually 60 MHz). Pentium wins over the bus, both for accessing the secondary cache, and for accessing the main memory and other attachments, for example, a set of PCI chips.
Using a specially designed bus for access to the secondary cache improves the performance of the computing system. In the first place, when it is possible to achieve complete synchronization of the hardware of the processor and the bus; in a different way, the competition with other operations of introduction and withdrawal is turned off, and it is connected with the cim blockage. The cache bus of the other peer is reattached to the external bus, through which access to the memory of that external extension is obtained. 64-bit zvnіshnіshnya bus can work with zі shvidkіstyu, schо lіvnієє halves, odnієї іnієї ї or іnіїї іnіїї ї іnіїї ї іnіїї іnіїїї ї іnіїї іnіїїї іnії іnієїї її іnіїїї vіdnostі protsessі, vіdkomі bus storіdnoj kasіu praciuє nezalezhnost' on vnіyі shvidkostі.
Combining a processor and a secondary cache in one package and linking them through a visible bus is a direct way to improve productivity, which is common in the most difficult RISC processors. So, the processor Alpha 21164 of the company "Digital" has a cache of another equal size of 96 kb of allocation in the processor core, like the first cache. This will ensure even higher cache productivity for the account by increasing the number of transistors on a crystal up to 9.3 million. The productivity of Alpha 21164 is 330 SPECint92 at a clock frequency of 300 MHz. The performance of the P6 is lower (according to "Intel" - 200 SPECint92 at a clock frequency of 133 MHz), however, the P6 provides the best speed/performance for its potential market.
Assessing the performance/productivity of the next step, if P6 can be more expensive than its competitors, most other processors can be honed by an additional set of memory chips and a cache controller. In addition, in order to achieve equal productivity, work with a cache, other processors will be responsible for the larger cache, lower than 256 kb to the world.
Intel, as a rule, promotes numerical variations of its processors. To work with the method of satisfying the different manpower of system designers and depriving less space for competitors' models. You can admit that it’s not a problem after the start of the release of P6 to be like a modification with an increase in the volume of the secondary cache memory, so it’s cheaper zovnіshnіm roztashuvannyam secondary cache, but for the saved seen bus between the secondary cache and the processor.
Pentium as point of reference
Pentium processor with its pipeline and superscalar the architecture of the reach of the enemy equal productivity. Pentium avenge two 5-stage pipelines, so they can process in parallel and beat two commands per machine cycle. At the same time, more than a couple of commands can be written in parallel, so that the program can go one by one and be satisfied with the sing rules, for example, the presence of register deposits for the type of record after reading.
P6 to increase the throughput capacity of buildings to move up to one 12-stage conveyor. Reduce the number of stages to change the skin stage of work and, as a last resort, change the hour of the change of the skin stage team to 33 cm against Pentium. Tse means that it is different with the development of P6 and technology, and with the development of a 100 MHz Pentium, it was developed before the introduction of the P6 with a clock frequency of 133 MHz.
Possibility of superscalar architecture Pentium, with її zdatnistyu up to two instructions per cycle, it would be important to overdo it without a new approach. Zastosovaniya at P6 new pіdkhіd usuvaє zhorstku zalezhnіst mіzh traditional phases "vybіrka" and "vikonannya", if the sequence of passage of commands through the cycle of two phases depends on the sequence of commands in the program.
A new approach to using the so-called pool of commands and new effective methods for transferring the future behavior of the program. In this case, the traditional phase of "visiting" is replaced by two: "dispatching/visiting" and "dispatching". As a result, the teams can begin to win in good order, or complete their win in the same order in the program. The P6 core is implemented as three independent attachments that interact through the command pool (Fig. 1).
The main problem on the way to increase productivity
The decision about the organization of P6 as three independent and interoperable through the pool of commands was adopted after a real analysis of factors, which intersect the productivity of modern microprocessors. The fundamental fact, which is true for the Pentium and the richness of other processors, is true for the fact that, in the case of real programs, the processor's fatigue does not win all over the world.
In that hour, the speed of processors for the remaining 10 years grew less by 10 times, the hour of access to the main memory changed by only 60 times. The cost of increasing the speed of the robot from memory according to the date of the speed of the processor was the same fundamental problem, which happened to be broken during the design of the P6.
One of the possible approaches to the solution of the problem is to transfer the center of gravity to the development of highly productive components, which will make the processor sick. However, the mass production of systems, which include a high-performance processor, and high-specialized microcircuits, is too expensive.
You could try to solve the problem with the help of brute force, and at the same time increase the size of the cache of another equal, in order to change the number of drops in the amount of necessary data from the cache.
The decision is more efficient, but it is also over the top expensive, especially looking at the current swedish vimogi to the components of the cache of another equal. P6 projected from the point of view of the effective implementation of the whole computing system, and it seemed that the high productivity of the system as a whole could be reached with the most cheap memory subsystems.
in such a manner, a combination of such architectural methods is implemented in P6, such as shortening the transfer of transitions (maybe the next sequence of commands is correctly determined), analysis of data flows (the optimal order of the sequence of commands is determined) and viperegialization of the transition (the sequence of commands is transferred and the optimal order is allowed to be allowed without simple productivity). according to the introduction to Pentium with different technologies of production. This combination of methods is called dynamic vikonann.
In this hour, "Intel" is developing a new 0.35 micron chip technology that will allow the release of the P6 processor with a core clock speed of over 200 MHz.
P6 is a platform for encouraging hard servers
Among the most significant the trend in the development of computers can be called the remaining fates, as the ever-growing development of systems with the improvement of processors of the x86 family as servers of add-ons, and the growing role of "Intel" as a supplier of non-processor technologies, as well as from the bus, mesh technologies, video compression, flash memory and system administration.
The release of the P6 processor continues the policy of transferring the possibilities, which previously allowed only more expensive computers, to the mass market. For the internal registers of P6 transfers, parity control, and the processor core, which connects, and the cache of another equal 64-bit bus, is equipped with the tools for displaying and correcting pardons. Due to the new possibilities of diagnostics in P6, we allow the breeders to design new systems. In P6, the ability to withdraw through the contacts of the processor is transferred, or for help software information about more or less 100 replacement processors or other ones that appear in the new, such as the presence of data in the cache, instead of registers, the appearance of self-identifying code, and so on. The operating system and other programs can read the information assigned to the processor. P6 also implements a polling of control points, so that it is possible to get a computer in the early fixation of a camp at the time of a pardon.
Similar documents
The cost of counting technique has been called out for a long time; Burkhlivy development of numerical technique. The creation of the first PCs, mini-computers, starting from the 80s of the XX century.
abstract, additions 25.09.2008
Characteristics of the systems of technical and preventive maintenance of the accounting equipment. Diagnostic programs for operating systems. Interaction of automated control systems. Defend the computer in the form of unfriendly intrusions.
abstract, additions 25.03.2015
Development of an information-analytical system and analysis and optimization of changes in the calculation technique. The structure of automated chirping by means of computational techniques. Software security, cost-effectiveness of the project.
thesis work, donations 05/20/2013
Manual stage development of counting technique. Positional number system. Development of the mechanics of the XVII century. Electromechanical stage of development of numerical technology. Computers of the fifth generation. Parameters noteworthy features supercomputer.
course work, donations 04/18/2012
Attached is the principle of a robotic personal computer (PC). Diagnosis of PC performance and indication of faults. Head of technical service for the accounting equipment. Development of techniques for improving technology at the manufacturing station.
course work, donations 07/13/2011
The development of foreign, domestic practice in the development of numerical technology, and the prospects for the development of EOM in the next hour. Technology of using computers. Stages of the development of the numbering industry, in our country. Zlittya PC and made a call.
course work, donations 04/27/2013
Classification of design procedures. History of the synthesis of computational technology and engineering design. Functions of computer-aided design systems, their software security. Features of the logging of trivi- mer scanners, manipulators and printers.
abstract, additions 12/25/2012
Data processing automation. Informatics and її practical results. The history of the creation of digital enumeration techniques. Electromechanical calculation machines. Selection of electronic lamps and EOM of the first, third and fourth generations.
thesis work, donations 06/23/2009
Understand that characteristic of a personal computer, its main parts and its recognition. Learn the science of informatics and especially the organization of work at the office of computational technology. The possession of working areas and the stagnation of software security.
abstract, additions 09.07.2012
Warehouse of the accounting system - computer configuration, hardware and software. Attachments that adjust the hardware configuration of a personal computer. Main memory, I/O port, peripheral adapter.
Counting technique
Basic understanding.
Often the concept of "calculating technology" is distinguished from the concept of "computer". At this point of view, under the understanding, they may be on the verge of the step:
Application: Computer(Engl. computer- "calculation") - the machine should be calculated.
For help, calculate the building computer and process the information for a predetermined algorithm. In addition, more computers save information and search for information, display information on see different pristroїv vdachі іnformatsії The name of the computer was taken away for its main function - the calculation. In short, the main functions of computers are the processing of information and management.
Basic principles: The computer placed in front of him can take care of the additional movement of any mechanical particles, flurry of electron flows, photons, quantum particles or a variety of effects in any other kind of good physical phenomena.
The widest range of computers has been given the title of "electronic counting machine", EOM. Vlasne, for the most important people, the words “electronic counting machines” and “computers” have become synonymous words, even though it’s not true. The largest expansion type of computer is the personal electronic computer.
The architecture of computers can directly model the solution of the problem, as close as possible (with a sensible mathematical description) to the physical reality. So, electron flows can be victorious as a model of water flows when modeling dams or rowing. Analogue computers were constructed in a similar manner in the 60s of the XX century, but the past years began to become a rare phenomenon.
In most modern computers, the problem is usually described in mathematical terms, in which all the necessary information is given in two form (it looks like one and zero), after which the processing can be reduced to a simple logic algebra. Oskіlki practically all mathematics can be reduced to the level of boolean operations, do it electronic computer can be used to complete more mathematical tasks (as well as more tasks for processing information, which can be easily reduced to mathematical ones).
They showed that computers can still be virish, not be a mathematical problem. Formerly, the task, which cannot be done with help of computers, was described by the English mathematician Alan Turing.
The result of the above task can be presented in the form of additional information input and output devices, such as lamp indicators, monitors, printers, etc.
Coristuvachi, scho start, and children often forcefully accept the idea that a computer is just a machine and cannot independently "think" or "understand" those words, like wines show. The computer is no longer mechanically modifying the given program lines and colors for the additional additions to the input and output. The human brain itself recognizes in the image on the screen image, the numbers of the words and the words give them those other meanings.
At a glance, I came across informatics as a science deacon, to talk about science as “calculating technology”.
Designated: Informatics and calculation technology- This is the area of science and technology, as it includes a collection of tools, methods and methods of human activity, directing to creation and stosuvannya:
· EOM, systems that measure;
· Automated systems for processing information and management;
· Computer-aided design systems;
· Software maintenance of computational equipment and automation systems.
Designation: Counting technique - ce
1) the field of technology, which involves the automation of mathematical calculations and processing of information in various fields of human activity;
2) the science of the principles of encouragement, di ї that design of tsikh koshtіv.
§2. "Counting technology" = "computer".
Etymology
Word computerє similar to English words to compute, computer, yakі are shifted like “calculate”, “calculate” (the English word, in its own way, resemble the Latin computo-"count"). Spoiler in English the whole word meant a person, as if arithmetic calculations were made from irradiated or without irradiated mechanical outbuildings. Nadalі yogo meaning was transferred by the machines themselves, the prote of modern computers wins an impersonal task, not connected directly with mathematics.
First interpretation of the word computer appeared in 1897 in the Oxford English Dictionary. Yogo stowage todі rozumіl kom'yuter as mekhanіchnіchnіchnіchnіchnіchnіvnіy pristrіy. In 1946, the vocabulary was replenished with additions, which allow to expand the understanding of digital, analog and electronic computers.
Classification of computers
There are no clear between classes of computers. In the world of advanced structures and technologies of variation, new classes of computers are being introduced, between the main classes are constantly changing.
Іsnuyut different classifications of computer technology:
I. following the principle of
1. analogues(AVM),
2. digital(TsVM)
3. hybrids(GVM)
ІІ. for the stages of creation (for generations)
1. 1st generation, 50 pp.: EOM on electronic vacuum tubes;
2. 2nd generation, 60 pp.
3. 3rd generation, 70 pp. Note. An integrated circuit is an electronic circuit of a special purpose, which looks like a single conductor crystal, which is similar to a large number of diodes and transistors.
4. 4th generation, 80 pp ..: EOM on large and super-large integrated circuits - microprocessors (tens of thousands - millions of transistors in one crystal);
5. 5th generation, 90 pp ..: EOM with dozens of parallel microprocessors that allow efficient knowledge processing systems; EOM on overfolding microprocessors with a parallel-vector structure, which chirp dozens of the last program commands at once;
6. 6th and next generation: optoelectronic EOM with massive parallelism and neural structure - with a divided array of a great number (tens of thousands) of clumsy microprocessors that model the architecture of neural biological systems.
ІІІ. by confession
1. universal(infamous confession),
2. problem-oriented
3. specialization
1. Basic EOM .
2. Universal EOM.
3. Specialized by EOM .
1) Keruyucha EOM .
2) Onboard EOM .
3) Seen by EOM .
4) Pobutova (home) EOM .
IV. beyond the borders functionality
1. oversized (superEOM),
2. great,
4. nadmalі (microEOM)
1) universal
a) rozrakhovanі on rich coristuvachіv
b) one-column (personal)
2) specialization
a) insurance on a rich coristuvachiv (server)
b) one-column (working stations)
v. According to the minds of exploitation, computers are divided into two types:
1. office (universal);
2. special.
Slid indicate what is the other classification. For example:
· Architecture.
· For productivity.
· For the number of processors.
· For calm authorities.
Short description class of computers
Following the principle of diy
Criteria for rozpodіlu counting cars on qi three classes є the form of presentation of information, for which stench is practiced (marveling at little ones).
Rice. Two forms of given information in cars:
a - analog; b - digital pulse.
Digital enumeration machines (TsVM)- Calculation of discrete machines, processing information presented in discrete, or rather, in digital form.
Such counting machines are often called EOM (electronic counting machines, electronic counting machines). The most widely used digital computer with electrical data of discrete information - electronic digital computing machines. , call simply electronic counting machines(EOM), without riddles about their digital character.
On the view of AVM, EOM, the numbers are presented in the sequence of numbers. In modern EOM, numbers are represented as codes of two equivalents, so as combinations of 1 and 0. In EOM, the principle program management. EOM can be subdivided on digital, electrified and analytical (perforating) enumeration machines.
EOM is subdivided into large EOM, mini-EOM and microEOM. The stinks are inspired by their architecture, technical, operational and overall characteristics, areas of congestion.
Advantages of EOM:
§ high accuracy of calculation;
§ universality;
§ automatic input of information required for the completion of the task;
§ raznomanіtnіst zavdan, yakі vyrіshyuyutsya EOM;
§ nezalezhnіst kіlkostі obladnannya vіd skladnostі zavdannya.
EOM shortfalls:
§ foldability of preparation of tasks until completion (needs special knowledge of methods for completing tasks and programming);
§ lack of knowledge of the process interruption, foldability of change of parameters of these processes;
§ folded structure of EOM, operation and technical maintenance;
§ help of special equipment for an hour of work with elements of real equipment
Analogue enumeration machines (AVM)- Counting machines without interruption, working with information presented in a non-interruptible (analogue) form, that is. as uninterruptedly low is the value of any physical quantity (mostly of the electrical voltage).
Analogue calculating machines are also simple and easy to operate; programming for the cherry on them, ring out, not a labor worker; The speed of task execution changes according to the operator’s requirements and it can be reduced to a large scale (bigger, lower for a digital computer), but the accuracy of task execution is even low (a noticeable error of 2-5%). On the AVM, the most efficient way to solve mathematical problems is to achieve differential equality, which does not require folding logic.
This is the calculation machine without interruption, as it processes analog data. Won is assigned for the implementation of the spewing spіvvіdnoshenie between physical quantities, which change without interruption. The main halls of zastosuvannya po'yazanі z modeling of various processes and systems.
In AVM, all mathematical quantities appear as uninterrupted values of any physical quantities. Head rank, like a machine change, the voltage of an electric lance is acting. These changes are subject to the very laws that change the assignments of functions. In these machines, the method of mathematical modeling is created (the model of the object is created, which is being carried out). The results of the solution are displayed at the sight of the electric voltage at the function for an hour on the oscilloscope screen or are fixed by the simulation devices. The main features of AVM are the difference between linear and differentiating lines.
AVM advantages:
§ high speed rozv'yazannya tasks, por_vnyanna zі shvidk_styu passage of an electric signal;
§ simplicity of AVM design;
§ ease of preparatory work before opening;
§ the accuracy of the transfer of the subsequent processes, the possibility of changing the parameters of the subsequent processes at the next hour of the extension.
AVM shortcomings:
§ low accuracy of the obtained results (up to 10%);
§ algorithmic exchange of problems;
§ manual entry of the rozvyazuvannya into the car;
§ great obligation of the backyard possession
Hybrid counting machines(GVM) - counting machines combined dії, pratsyyut s іnformatsiєyu, presented in digital, і analog formі; stinks to eat at your own expense of AVM and CVM. GVM dotsіlno vikoristovuvaty for vyvіshennya zavdan adminіnnya foldable shvidkodіyuchmi technical complexes.
One-of-a-kind machines are called combined counting machines», « analog-digital counting machines (ATsVM)»
The stench may have such characteristics, like a code, simplicity of programming and versatility. The main operation is integration, as if it were for the help of digital integrators.
For such machines, the numbers are represented as yak і EOM (sequence of digits), and the method of execution is yak і AVM (method of mathematical modeling).
Behind the stages of creation
Podіl kom'yuternoї tehn_n_nіnya - duzhe umovna, nesuvora klassifіkatsіya counting systems equal to the development of hardware and software zasobіv, and navit ways of interfacing with a computer.
The idea of extending the car for a generation of viklikan life tim, in an hour short stories computer technology developed its own development of the great evolution like a sensi element base(Lamps, transistors, microcircuits and etc.), so in the sensational changes in the structure, the appearance of new possibilities, the expansion of the areas of congestion and the nature of the vikoristannya.
EOM of the 1st generation[ first generation computer ]
Until the first generation, cars, created between the 50s, sounded.
All EOMs of the 1st generation were broken based on electronic lamps what robbed them of inappropriateness - the lamps had to be changed often.
Rice. electronic lamp
qi computers were majestic, non-handling and overly expensive machines, as if only great corporations and orders could come. The lamps saved the majestic amount of electricity and saw a lot of heat.
The set of commands is small, the scheme of arithmetic-logical add-on and add-on keruvannya is simple, the software was practically not bulo. Indicators obsyagu operative memory and security codes were low. For introduction-vivodu punched lines, punched cards, magnetic lines and other attachments were made.
Rice. Card
The speed code is close to 10-20 thousand operations per second.
But the less technical side. Another important thing is the way of using computers, the style of programming, especially the mathematical security.
The programming was done with the help of low level programming language. The programs of these machines were written my particular car. A mathematician who wrote the program, sitting at the remote control of the machine, entering it, adjusting the programs and vibrating the rahunok behind them. The process of nagodzhennya buv found in an hour.
Irrespective of the availability of opportunities, these machines allowed the most convenient rozrahunki, necessary for forecasting the weather, the accomplishment of the tasks of nuclear energy and other.
Dosvіd vykoristannya machines of the first generation showing what is really great razryv mizh hour, scho stained for rozrobka programs, that hour rahunku.
Rice. a - Computer "Eniak", b - EOM "Ural"
These problems began to be done as a way of intensive development of the automation of programming, the creation of service software systems, which make it easier for the robot to work on the machine and increase efficiency and complexity. Tse, with his calling, longed for significant changes in the structure of computers, investments in those who wanted to bring them closer to the vim, yakі z dosvidu exploitation of computers.
First generation machines: MESM (small electronic machine), BESM, Strila, Ural, M-20.
EOM II-th generation[second generation computer ]
The machines of this generation were constructed around 1955-65.
AT 1958 r. in the EOM (JOM of the 2nd generation) were stagnant conductor transistors, wine 1948 William Shockley.
History of winemaking:
· On June 1, 1948, on one of the sides of the New York Times, devoted to radio and television broadcasting, there was a modest announcement about those that Bell Telephone Laboratories had built an electronic appliance, a building replacement of an electric lamp. Theoretical physicist John Bardeen and the firm's pioneering experimenter Walter Brattain created the first transistor that did. The chain was a point-contact attachment, in which two metal "antennae" were in contact with a bar of polycrystalline germanium.
· The creation of the transistor was re-engineered, may be a 10-row robot, as in 1938 and 1939 the physicist-theorist William Shockley rozpochav. Vtіm, more precisely, the history of the transistor began richly earlier. Already in 1906 the Frenchman Picard propagated a crystal detector, then in 1922 the radiophysicist O.V. Losev, having shown the possibility of strengthening that generation of colivans with the help of such detectors. After 3 years, Professor of the University of Leipzig Julius Lilienfeld tried to create a subdued navprovodnikovy device. The prote of the experiment was forgotten. They guessed about them only after that, like a transistor, having won all-world recognition.
· It became tse, before the speech, finish the shvidko. PISLAL Killykhhkhhhhkiv technologists of the technology of napivprov Prudnikovye, the guilty of the Novykh structural (Zokrem, the Fasten Transistor, the patented U. Shoklі in 1951 Rotsi) the Low of the American viophah of the transistor, the Yaki spokes in the main
The transistors were bigger, more advanced, smaller, they could significantly reduce the number of calculations, they were small for a large operational memory. 1 building transistor replace ~ 40 electric lamps and work with more security.
At every turn, these computers zastosovalis like electronic lamps, and discrete transistor logic elements. Pіznіshe discrete transistor logic elements.
§ In the spirit carrying information victorious magnetic lines("BESM-6", "Minsk-2", "Ural-14") and magnetic cores.
§ Їх operational memory bula prompted on magnetic cores.
§ Having expanded the range of I/O, which zastosovuetsya, appeared highly productive attachments for robots with magnetic stitches, magnetic drums and first magnetic disks.
§ In the spirit software began to win movie programming of a high level. Allow such language descriptions of all the necessary sequence of enumerations at a clear, easily grasped look. The program, written in an algorithmic language, did not understand the computer, which takes less than its own commands. Therefore, special programs, as they are called translators, shifting the program from a high level language to a machine language.
§ Showing up a wide range of library programs on the accomplishment of various mathematical problems.
§ Z'appeared monitor systems that control the broadcast mode and program. From monitor systems, modern operating systems have grown. Otzhe, operating system software extensions attach a computer. For some machines of another generation, operating systems with limited capabilities have already been created.
§ Cars of another generation bula domineering software inconsistency, as it facilitated the organization of the great information systems. Therefore, in the mid-60s, there was a transition to the creation of computers, programmatically smart and prompted on a microelectronic technological base.
§ Shvidkodiya- Up to hundreds of thousands of operations per second.
§ Memory capacity- Up to a few tens of thousands of words.
Particularly, for the first generation.
1. Higher arrogance.
2. Less energy saving.
3. More swedkodiya for rahunok:
· Increased speed of shifting of rachunks and memory elements
· Changes in the structure of machines.
Rice. a - Transistor, b - memory on magnetic cores
Starting from another generation, the machines began to grow large, medium and small after the signs of expansion, varnosti, and number of possibilities. So, small machines of another generation (“Nairi”, “Razdan”, “Mir” etc.) with productivity close to 10 4 operations per second were similar to the 60s as a whole available to a skin university, just like the BESM -6 was small professional displays (і varіst) are 2-3 orders of magnitude higher.
Rice. BESM-6.
EOM III-th generation[third generation computer]
Have 1960 r. The first integrated circuits (IC) appeared, as if they had a wide width at the link with small sizes, but great possibilities.
Rice. Integrated circuits
ІС (integrated circuit) - silicon crystal, the area of which is approximately 10 mm2. The first IC is designed to replace tens of thousands of transistors. One crystal beats the same robot, like a 30-ton Eniak. And a computer made of ICs can achieve productivity of 10 million operations per second.
In 1964, the rotating company IBM announced the creation of six models of the IBM 360 (System 360) family, which became the first computers of the third generation.
Machines of the third generation - tse families of machines from a single architecture, that is. software crazy. As an elemental base, such vicors have integrated circuits, which are also called microcircuits.
Machines of the third generation may have a different operating system. The stink of the possibility of multiprogramming, tobto. one-hour vikonanny kіlkoh prog. The management of the memory is rich, the outbuildings that resources began to take over the operating system or the machine itself.
Apply machines of the third generation - the IBM-360, IBM-370, EC EOM (Single EOM System), SM EOM (Small EOM Family) and others. The code of machines used in the middle of the family changes from a few dozen thousand to a million operations per second. The operational memory capacity is hundreds of thousands of words.
First integrated circuits (IC)
The first integrated circuit, developed in 1960, was the prototype of today's microchips. The integrated circuit is made up of miniature transistors and other elements that are mounted on a silicon crystal.
37 years ago, in 1964, IBM announced the creation of six models of the IBM 360 family (System 360), which became the first computers of the third generation.
Models are small єdinu system of commands and vіdіznyalis one vіd one obsyagy operative memory and productivity. Thomas Watson, Jr., head of IBM, called the emergence of this family of machines "the most important development in the history of the company." The first machines of the IBM 360 series were brought to replacements in the other half of 1965, and until 1970 the company produced about 20 models, but none of them were brought to serial production.
With the creation of models of the vicorist family, a number of new principles worked, which made the machines universal and allowed, with the same efficiency, to set them up for the completion of tasks in various fields of science and technology, and for the processing of data in the sphere of management and business in the business (number 360 machines are practicable at all straight lines - no more than 360 °). The most important innovations were:
· Elementary and technological base of machines of the third generation;
· Program summіsnіst vsіh models of the family;
· an operating system that could replace translators for the most extensive programming languages at that time (Fortran, Cobol, RPG, Algol 60, PL/1), and it was possible to include translators for other languages in the system;
· "Universality" of the command system, as a way to add additional commands for different purposes to the so-called standard command system;
· the possibility of connecting a large number outbuildings standard spoluchennya tsikh pristroїv with the processor through the equipment kanalіv vyazka (when it was possible to unite a sprat of machines into one calculation system);
· Organization of memory, so as not to lie in the physical implementation, which will ensure the simple movement of the program;
· A powerful system of hardware and software interruptions, which allowed organizing the efficient operation of machines in real time. The creation of models of the IBM 360 series made a significant contribution to the entire development of computer technology. The structure and architecture of these machines with the same number of changes in the element base were implemented in a number of families of the EOM of rich countries.
EOM ІІІ-th generation. AT 1960 r. first came integrated circuits (IC), yakі nabuli wide width at the link with small sizes, but majestic capacities.
· Computer with IC accessibility productivity in 10 million operations per second.
· In 1964, IBM announced the creation of six models of the IBM 360 (System 360) family, which became the first computers of the third generation.
· Machines of the third generation families of machines from a single architecture, then. programmatically crazy.
· In the pit element base they have a twist integrated circuits, which are also called microcircuits.
· Third generation machines operating systems.
stink multiprogramming capabilities, then. one-hour vikonanny kіlkoh prog.
EOM IV generation[fourth-generation computer]
On the cob of the 1970s, mid-integrated circuits began to be victorious. And then - great integrated circuits.
Crimea changes element-technological bases, new ideas about the structure of counting machines, programming, debugging and exploitation of counting systems appear.
First, large integrated circuits (ВІС) began to zastosovuvatsya, as if for tension approximately 1000 ІС. This was caused by a decrease in the number of computers. Have 1980 r. the central processor, a small EOM, appeared to be able to be placed on a 1/4 inch (0.635 cm 2) crystal.
BICs were already stoked on such computers, like “Illiak”, “Elbrus”, “Makintosh”. The speed code of such machines can store thousands of millions of operations per second. The capacity of RAM (working memory) grew up to 500 million double discharges. In such machines, a number of commands over a number of sets of operands are concatenated at once.
Three points of the structure's view: machines of the second generation rich processorі bagatomachine complexes, which is used for the solemn memory of that solemn field of outbuildings. The size of the operational memory is about 1 - 64 MB.
The expansion of personal computers before the end of the 70s led to a significant reduction in the consumption of large EOM and mini-EOM. It was the subject of serious concern for IBM (International Business Machines Corporation) - a leading company for the manufacture of great EOMs, and in 1979. IBM has tried its hand at the personal computer market by creating the first personal computer, the IBM PC.
Їх characteristic:
· Zastosuvannya personal computers;
· Telecommunication processing of data;
· computer networks;
· Widespread zastosuvannya database management systems;
· Elements of intellectual behavior of data processing systems and attachments.
EOM of the fourth generation - vicorist large and super large integrated circuits(ВІС and НВІС), virtual memory , a rich processing principle with parallel operations dialogue.
EOM V-th generation[fourth-generation computer], EOM of the VIth generation and so far
EOM of the fifth generation - 90th pp. EOM on overfolding microprocessors with a parallel-vector structure, which can churn out dozens of the last program commands at once.
EOM of the sixth generation and the next generation: optoelectronic EOM with massive parallelism and neural structure - with a divided network of a large number (tens of thousands) of clumsy microprocessors that model the architecture of neural biological systems.
Skin advance of the generation of EOM best features. Thus, the productivity of the EOM and the capacity of all outbuildings increase, as a rule, by an order of magnitude more.
Just as before the EOM retailers from I to IV generations there were such tasks, as an increase in productivity in the field of numerical surveys, reaching a great memory capacity, then the main tasks of the EOM retailers of the fifth generation (advancing) were the creation of piecewise machine intelligence (possibility of given facts), the development of "intellectualization" of computers - usunennya bar'er between people and computers. Computers will be able to process information from a handwritten or handwritten text, from blanks, from a human voice, to recognize the voice behind the voice, and to translate the translation in one language. Tse allow us to speak with EOM to us coristuvachas, navit tim, who do not have special knowledge in this gallery. EOM will be a helper for people in the mustaches.
The development of upcoming generations of computers is carried out on the basis of grand integrated circuits of the advanced level of integration, based on optoelectronic principles ( laser, holography).
There will be a clear transition from the processing of data to acquire knowledge.
It is believed that the architecture of the computers of the future generation contains two main blocks. One of them is traditional computer. But now I can’t get in touch with the koristuvach. Tsey zv'yazok zdijsnyuє block, titles termin "intellectual interface". Yogo zavdannya - understand the text, writing with my natural language and how to avenge my mind zavdannya, and translate yoga into a working program for a computer.
Also, the problem of decentralization is growing. computer mesh like the great ones, which are located on a significant vіdstanі one type of one, so are the miniature computers, placed on one crystal of the heater.
By confession
Universal EOM is recognized for the implementation of the most sophisticated engineering and technical tasks: economic, mathematical, informational and other tasks, which are influenced by the complexity of algorithms and the great obligation to obtain data. The stench is widely vikoristovuyutsya in the centers of collective corystuvannya and other straining obschislyuvalnye complexes.
Characteristic rices of universal EOM є:
- high productivity;
- diversity of forms of data collection: double, tenth, symbolic, with a great range of changes and high accuracy of their data;
- the nomenclature of victorious operations is great, such as arithmetic, logical, and special;
- large capacity of operational memory;
- the organization of the system of input and output of information has been sanctioned, which ensures the connection of various types of outbuildings.
Problem-oriented EOM serve for the accomplishment of a larger number of tasks, tied, as a rule, with technological objects; registration, the accumulation of such a small collection of data; vikonannyam rozrahunkіv schodo vydnosno clumsy algorithmіv; stench can be interchanged with universal hardware and software resources.
Prior to the problem-oriented EOM, one can see, zocrema, various control and calculation complexes.
Specializations EOM are selected for the implementation of the narrow stake of the task or the implementation of the singing group of functions. Such a university orientation of EOM allows to clearly specialize its structure, to significantly reduce its folding and versatility while saving high productivity and reliability of its work.
Prior to the specialization of EOM, one can, for example, program microprocessors of a special purpose; adapters and controllers that enable logical functions to control a number of clumsy technical outbuildings, units and processes; add-on for the use of robotic vuzlіv billing systems.
Basic EOM [original computer] - EOM, which is a cob vent model in the EOM series of a single type or mind.
Universal EOM[universal computer] - EOM, is recognized as a high class leader. EOM of this class can be decomposed and algorithmically the same system of operations, hierarchical structure memory that broken system outbuilding danich.
Specialized by EOM [specialized computer] - EOM, recognized for the highest class of singing tasks. Characteristics architecture machines of this class are distinguished by the specifics of the plant, on the basis of the stench they are oriented, which makes it more efficient to universal EOM. Before the category of specialists, there can be butts, zokrema, - "ker_vniki", "boards", "buttovs" and "sights" of the EOM.
Keruyucha EOM[control computer] - EOM, recognized for automatic keruvannya object (attachment, system, process) in real time. Every EOM with the object of management is carried out for help analog-digitalі digital-analog converters.
Methods of organizing software and technical facilities in the AWP complexes may be prominent in the context of the analysis of the processes of operational management of the production (OUP) of industrial enterprises, the purpose of which is to minimize the use of all types of resources in the production of nomenclature.
Methods of organizing software and technical facilities in the AWP complexes may be prominent in the context of the analysis of the processes of operational management of the production (OUP) of industrial enterprises, the purpose of which is to minimize the use of all types of resources in the production of nomenclature.
Synthesis of methods and models of organization of programmatic and technical feasibility when presented by AS OUP as a complex of AWPs of state transparent recruitment teams can go through two stages: the stage of assigning a rational warehouse of BT and the stage of solving the problem of distribution of resources for the calculation of the chemical systems.
The technical (equipment) sum of new BT gains is due to the transfer to the BT park at the deputy’s owner and to the BT park, which is predicted for the arrival of the nadal. Practice shows that this show is one of the most important, which is taken to the point of respect when choosing BT. The tendency to add money to the VT of equipment is sumptuous with obvious reasons due to a variety of objective and subjective reasons, where there is no place left for borrowing and the psychology of the manager, which is almost indebted to the success of the success of the success of this class of equipment. Software confusion what depends on the sum of the hardware-implemented system of commands, the sum of the data presentation formats, the sum of translators, DBMS, etc. Significant contribution of this indicator to the resource consumption can be explained by the great achievements in the early preparation of normative, archival and statistical data, as well as by the specialization of the prepared personnel at the enterprise, which can be achieved with specific basic software security.
Exploitation summ_snіst seredinі pribynі complex zabіv VT, scho allows razі vyhod out of tune okremy modules Workstation either promptly replace the module, which is in order, or carry out reassignment of vicoristing outbuildings between specific workstations in the boundaries of calculating the resources of all complexes (in the middle of the workshop complex, in the middle of the inter-workshop complex, in the middle of the system of any enterprise).
Reliability of VT protection for technical minds and її resilience to specific minds of the work: vibration, oxidation, saw, gas contamination, strains are too thin. require additional support for the defense.
Sukupna svydkіst vyvіshennya vyvіshennya funkіshennya zavdanі for types of workstation complex - svydіkіstі obrobki іsnuyuchih obyagіv dannyh і іrіnіh modes of work. It is not enough to know the exact information base of a specific workstation and passport characteristics and enumeration resources.
Therefore, for an approximate (ordinal) assessment of the value of the indicator, it is worth it, either to prove the operation on objects close to the class of VT, or the results taken on the simulation models, debasing the data from the obligatory data are real. Approximation of data, taken on control butts, can lead to a loss of results, which are reviewed by an order of magnitude in terms of taking away real estimates for an hour of operation of the system. Dzherelom's faults are most often the ambiguity of robotic algorithms, operating system utilities, communication protocols, drivers, and basic techniques in the operation of systems in a richly complex rich-tasking mode on the boundary resources of computational systems or for their bulk elements. In this way, the possibility of direct expansion with different characteristics of firmware processors, internal machine channels, links, channels, access to data for types of external extensions cannot be ineffective. In this hour, the number of rich processors and the realizable targeting of their current benefits do not allow to secure all the potential of the impersonal task of the control system of the PPP with the necessary accuracy. Therefore, it is necessary to carry out the details for the classes of the task of specific types of workstations with reference to a well-defined combination of tools for BT and basic software security.
The variability of the implementation of the “friendly interface” includes the programs for starting and the possibility of removing improvements from the work process on the workstation about ways to continue or complete the dialogue.
Possibility of changing the warehouse and changing the functions implemented on specific workstations, redesigning the system between staff.
Security could protect against unauthorized access to databases of knowledge and databases, as well as security of their “foresight” at times of consumption.
Not long ago, the term "calculating technique" appeared in the future. This understanding of the heart is not small enough for the consideration of all these aspects, as they are invested in a new day. And, unfortunately, more people care about the fact that computers and the numbering technique are synonymous words. Tse is clearly a pardon.
Counting technique: the meaning of the word
It is possible to interpret the meaning of this term in a different way, and thus more different vocabularies of cloudiness can be used in different interpretations.
ONE Yakshcho Pіdіita to Panitan Yak Bi Yakimos Uzagalnnems, can be smiled, he is computed by Technika - Tekhnichny, an attachment of mathematical posts, Priyomiv -Iaki otherwise. another phenomenon (physical, mechanical, then).
tse scho such a wide mind?
It’s been a long time since the number of technіka in the house of people has been completed. The most primitive outbuildings, which appeared hundreds of years before ours, can be called, for example, the very Chinese rahunki or the Roman abacus. Already in the other half of the last 1000 years, such extensions appeared, such as the Knepper scale, the Shikkard arithmometer, the calculus, and so on.
The interpretation of this term has gained more extended meaning with the appearance of the first EOM. It happened in 1946, when the first EOM was created in the USA, which was designated by the abbreviation ENIAC (in the SRSR, such an attachment was created in 1950 and the name MESM was not enough).
On this day, the interpretation has expanded even more. In this way, at the current stage, the development of technologies can be determined, what is the number of techniques - tse:
- computer system and the ability to manage the lines;
- automated control systems and data processing (information);
- automated designing, modeling and forecasting;
- software development systems and int.
Bones for calculating
Now let's wonder what the numbering technique is. The basis of any process is information or, as usual, say at once, data. But understanding information is important to make it subjective, shards for one person, such a process can carry a sense of purpose, but for others - not. In such a rank, for the unification of these letters of division, some kind of machine will be used for the processing of these data most widely.
Among the facilities themselves, you can see technical attachments (processors, memory, input/output attachments) and software security, without which everything is completely visible. Here, it is important to note that the system has a number of characteristic features, for example, integrity, organization, engagement and interactivity. There are more so called counting complexes, which can be added to rich processor systems, which will ensure that promotion rіven productivity, inaccessible to standard single-processor systems. You can only talk about those that stench is the main reason for calculating the stench. Naturally, you can add here the methods that will be used to describe mathematically what is going on in the process, or you can take a long time.
Powering modern computers
Three of these terms can describe the work of modern computers. As it was appointed more, the stench will be in your own hardware and software parts, moreover, one without another is impossible to function.
In this rank, the current computer (calculating technique) is the totality technical outbuildings, which ensures the functioning of the software environment for the performance of the singing tasks, and navpaki (succession of programs for the work of the “hall”). The most correct is the first step, and not the other, even if it is possible, this set is needed for processing the input information and drawing the result.
(calculating technique) includes a few of the main components, without which any system cannot do. Here you can store motherboards, processors, hard disks, RAM, monitors, keyboards, mice, peripherals (printers, scanners, etc.), disk drives and others. At the software security plan, the first place is to install operating systems and drivers. AT operating systems practice application programs, and the drivers ensure the correct functioning of all "outside" outbuildings.
Dekilka of words about classification
Modern counting systems can be classified according to how many criteria:
- principle dії (digital, analog, hybrid);
- generation (stages of creation);
- recognition (problem-oriented, basic, on-butovy, visual, special, universal);
- capacity and size (super-large, super-small, one-or-one for rich coristuvachs);
- wash zastosuvannya (home, office, virobnichi);
- other signs (number of processors, architecture, productivity, less power).
As I understood, it is impossible to hold clear cordons among the designated classes. In principle, be it a subdivision of modern systems on a group, everything looks like a day to mind.
