Shulou(Shulou.com)06/01 Report--

Three major pieces of storage and server

From the point of view of the implementation of the underlying principle, the physical hardware nature of storage and server are three major components: the combined use of CPU, memory and Icano.

Second, the server's combined application of the three major pieces.

From a computer's point of view, a server is also a computer, but it is more complex and advanced than ordinary computers. The birth of the server is to deal with enterprise-level applications and make the work more collaborative. Therefore, compared with home computers, the most important thing is to be stable, secure and make fewer errors. The applications and databases running on the server are very important to the enterprise, so the performance and stability of server-level memory, hard disk, network card, CPU and so on will theoretically be higher than that of ordinary computers. So the server CPU, memory, hard disk, network card and other accessories have been upgraded to make it more secure and stable. However, there is no difference in the combined use of CPU, memory and Imax O devices. In high-end servers, due to the support of multi-channel CPU, memory and Imax O devices, the combination of the application mode will be more complex.

1. System bus combination phase

In the stage of system bus combination, the CPU chip, memory and I-hand O devices are connected through the system bus to realize the communication and cooperation between them, which realizes the most original combination application and the most basic computing function.

From the beginning of the architecture design of the server or computer, the combined use of CPU, memory and Imax O devices is the most rough and simple. It connects CPU, memory and Imax O devices through a system bus, and the communication of CPU, memory and Imax O devices must be transmitted and synchronized through the system bus. Because of the differences between devices, in this simple digital system. In order to ensure that all the internal hardware units can work together quickly, CPU architecture engineers have designed a set of clock signals to operate synchronously with the system. CPU determines the frequency at which other devices work, and ensures that all devices work in synchronous mode. In this mode, all devices are limited to a general clock frequency that all devices can withstand, that is, the so-called "interlock" effect, the barrel principle tells us, in this case, the overall performance of the system will be pulled down by the lowest frequency device.

The following figure shows the basic block diagram of the original computer system. It consists of CPU, memory (RAM) and a number of Icano devices. All of these are connected through the system bus. CPU executes instructions stored in memory. These instructions can read data from the memory or Iram O device, and can manipulate the data, as well as write the resulting data to the memory or Iram O device.

How to start a computer with this architecture for the first time, because there is no startup code in memory. In order to solve the startup problem of the computer directly connected to the system bus, the initial solution is to use a front panel (The front panel) to control the startup process of the computer, and then use the ROM mode to control the startup of the computer.

Front panel architecture design:

The front panel is designed by adding devices with indicators, toggle switches and buttons to a metal panel, which are connected to the bus through wires, and then use a row of lights to represent the contents of one or more hardware registers. and allows the contents to be read directly when the machine stops. When the computer starts for the first time, booting through the front panel usually requires a series of complex operations. First, the operator needs to manually enter the program instructions containing a series of short boot instructions using the toggle switch on the front panel. The first step is to set the "address" information, and input the address representation information data in binary form by setting the address switch. There will be special binary buttons on the front panel, such as DEC PDP-8 or MITS Altair 8800 binary numbers grouped into three on the front panel, and each set of lights or switches represents a single octal number between 0 and 7. Next, the operator will set the value of the address, pass the value switch, and then enter a value for that address. The operator will then set the starting address of the bootstrap and activate the "RUN" switch to start the execution program. The bootstrap program usually reads and loads the program from perforated paper tape and eventually runs the computer.

The front panel can control the register addresses of memory controllers, CPU registers and other Imax O devices, the user can set the register address to check or change, write new data to the register location, and reset, start and stop CPU. Most front panels provide a way to read and write individual CPU registers, and CPU executes the program step by step one instruction at a time when instructions are entered through the front panel.

The following image shows the physical front panel of System/360 Model 91 (image from Wikipedia)

In the late 1950s, the computer was a fairly large and expensive machine. It cost much more than luxury cars or even houses at the time, and a large front panel full of switches and lights relatively increased the cost of computers. When the computer is powered on, CPU is not yet running, and operators can use the front panel to read and write the contents of memory without CPU help. CPU starts running when the appropriate program is stored in memory and the operator sets the front panel switch to the RUN position. Most computers built before 1975 have such front panels.

Each time the computer is powered on, the operator must enter the bootstrap into memory, which will read some data from an Imax O device (such as a card reader, tape, or disk). The loaded data forms a Mini Program, and then loads more data until the entire operating system or application is loaded. Depending on the computer architecture at that time and the type of Iamp O equipment, these instructions were about dozens of instructions and had to be entered bit by bit every time the machine was powered on.

In order to solve the tedious problem of computer system startup, the memory needs to use non-volatile storage media to save the startup code data in memory and ensure that the startup instructions are still stored in memory in the shutdown state. When CPU starts, the computer begins to read and execute instructions from a clearly defined location in memory. From the late 1950s to the early 1970s, most computers used core memory to store in-memory data, which was non-volatile. That is, when we start the computer, the memory still contains what was stored the last time we used the machine.

After the system starts, in order to further solve the tedious problem of initialization and operation of the subsequent Imax O device on the front panel, some computer designers add logic to the CPU so that CPU can load the startup code from the Imax O device on its own without any instructions in memory. The operator only needs to press the correct button to choose from the correct device to automatically start the loading code and run the Icano device.

When the hardware cost of the computer becomes cheaper due to the progress of science and technology and the improvement of technology, the expensive, large and obviously unfriendly front panel has become an urgent problem to be solved. In addition, the storage characteristics of memory make the capacity of memory unable to meet the needs of the system, and the cost of expanding capacity will remain high. So architects designed ROM (read-only memory) and CRT (picture tube) to replace the function of the front panel, and use large-capacity memory components to put the programs necessary for computer operation into the ROM. In the late 1970s and early 1980s, with the exception of some early microcomputers that still had a front panel, the vast majority of microcomputers had been replaced with ROM and CRT.

The purpose of ROM is that when the computer is powered on, the CPU starts executing instructions from a clearly defined address that points to read-only memory (ROM), as shown in the following figure. The program in ROM controls the function that the computer starts.

The internal data of ROM is recorded in a special way in the factory during the manufacturing process of ROM, and the contents can only be read and not changed. Once burned, the user can only verify whether the written data is correct and cannot make any further changes. If you find an error in the data, you have to abandon it and order a new copy. ROM solves the tedious problem of starting the electronic computer on the front panel, which makes it much more convenient to start the computer, but it also has some disadvantages. Because ROM defines the register address and boot address information of the internal components of the computer, the function of the computer is determined not only by its hardware, but also by the program stored in its ROM. To put it simply, ROM limits the specific specifications of the computer.

The scope of software in ROM varies greatly from computer system to computer system. On some machines, ROM simply loads the first sector of the floppy disk into memory, and then loads the program that CPU jumps to memory to run. The program you just loaded will load the operating system from the floppy disk into memory. If there is no suitable system disk on the floppy disk, the computer will not start at all and will prompt that the startup file cannot be found. On the other hand, if ROM can load the boot sector of the hard disk into memory and run from this program, thereby loading the operating system on the hard disk. In many computers, programs in ROM determine what the computer can do and how to do it. Generally speaking, the memory size is usually very limited, so storing important programs (or operating systems) in ROM means that you can provide more memory space for applications or data, and you can also accelerate the startup speed of the operating system. Based on this design and concept, some computers embed very limited firmware in ROM. These computers have a large number of programs on ROM, including the operating system. Such computers burn the operating system and functional implementation in ROM.

Judging from the components of today's computer, each component that makes up the computer contains the basic components of the chip, memory and Istroke O core. The operating systems of most of the modules or components are burned in the ROM of the components themselves. Some components of ROM are independent, and some components, memory and chips are completely integrated in the same chip module.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.