staycomputer scienceMedium,random access(sometimes calledDirect access）Represents accessing an arbitrary component in a set of sequences at the same time.Otherwise, it is calledSequential accessThat is, more time is needed to access a remote component.The basic structure of random access memory can be divided into three parts: storage matrix, address decoder and read-write circuit.
The so-called "random access" means that when the message in the memory is read or written, the time required is independent of the location of the information.In contrast, when reading or writing information in a Sequential Access storage device, the time required is related to the location (such as tape).
2. Volatility
RAM cannot retain data when the power is turned off.If data needs to be saved, they must be written to a long-term storage device (such as a hard disk).The biggest difference between RAM and ROM is that the data saved on RAM will disappear automatically after power failure, while the ROM will not.
3. High access speed
Modern random access memory has the fastest write and read speed among almost all access devices, and the access delay is also insignificant compared with other storage devices involving mechanical operation.
4. Need to refresh
Modern random access memories rely on capacitors to store data.When the capacitor is fully charged, it represents 1 (binary), and when it is not charged, it represents 0.Since the capacitor has more or less leakage, if no special treatment is made, the data will gradually lose over time.Refresh refers to reading the state of the capacitor regularly, and then recharging the capacitor according to the original state to make up for the lost charge.The need to refresh just explains the volatility of RAM.
5. Sensitive to static electricity
Like other fine integrated circuits, RAM is very sensitive to the electrostatic charge of the environment.Static electricity will interfere with the charge of the capacitor in the memory, causing data loss, or even burning the circuit.Therefore, before touching the RAM, you should touch the metal grounding with your hands.
Random Access Memory Overview
Announce
edit
Random Access Memory (RAM) iscomputerThe most familiar type of memory.RAM is called "random storage" because you can directly access any storage unit as long as you know the address of the memory row and memory column where the unit is located.
In sharp contrast to RAM is sequential access memory (SAM).The data storage units in SAM are arranged in a linear order, so they can only be accessed in order (similar to cassette tapes).If the required data cannot be found at the current location, the next storage unit must be searched in turn until the required data is found.SAM is very suitable for buffer storage. Generally, the data in the cache is stored in the same order as the call order (the quality cache in the graphics card is a good example).RAM can access data in any order.[1]
Random Access Storage Infrastructure
Announce
edit
The basic structure of random access memory (RAM) can be divided into three parts: storage matrix, address decoder and read-write circuit. The following is divided into three parts:
1. Storage matrix:The storage matrix is used to store the code to be stored. Each storage unit in the matrix is numbered with a binary code for querying this unit.This binary code is called an address.
Internal structure of RAM
2. Address decoder:The decoder can translate the input address into a level signal to select the response unit in the storage matrix.The addressing mode is divided into single addressing and binary addressing.Univariate addressing is also called one-way decoding or word decoding, and the decoding line output is the word selection line, which is used to select all units of the accessed word;Binary addressing is also called bidirectional decoding. Binary addressing can access each unit. The decoding line output by the X address decoder is used as the line selection line for "line selection";The decoding line output by the Y address decoder is used as the column selection line for "column selection", and the unit selected at the same time as the row and column selection line is the accessed unit, which can be "written" or "read out".
3. Read/write circuit:The read-write circuit is the control part of RAM, which includes chip selection CS, read-write control R/W and data input readout amplifier. Chip selection CS is used to read and write only when the terminal is powered down. The read-write control R/W is used to read out the RAM when the R/W terminal is powered up.Write when the R/W terminal is powered down.The output stage circuit generally adopts a three state output or open collector output structure to expand the storage capacity. If it is an open collector output (i.e. OC output), it should be externally connected with a load resistor.
Random storage purpose
Announce
edit
SRAM: Static random access memory adopts multi transistor design. Usually, 4-6 transistors are used for each storage cell, but there is no capacitor.SRAM is mainly used for caching.
DRAM: Each memory cell in dynamic random access memory consists of paired transistors and capacitors, which need to be refreshed continuously.
FPM DRAM: Fast page mode dynamic random access memory is the earliest DRAM.During the whole process of bit positioning by the memory according to the row and column address, the FPM DRAM must be in the waiting state, and the next bit data can be processed only after the data is read.The maximum transfer rate to L2 cache is about 176MB per second.
EDO DRAM: extended data output dynamic random access memory can process the next bit of data without waiting for the whole process.As long as the address location of the previous bit of data is successful, EDO DRAM will start addressing the next bit of data.It is about 5% faster than FPM.The maximum transfer rate to L2 cache is about 264MB per second.
SDRAM: Synchronous dynamic random access memory uses the concept of burst mode to greatly improve performance.This mode first locks a memory row when reading data, then quickly scans each memory column while reading the bit data on the column.The reason for this design idea is that most of the time the data requested by the CPU is adjacent in memory.SDRAM is about 5% faster than EDO RAM, and has become the most widely used desktop memory.The maximum transfer rate to L2 cache is about 528 MB per second.
DDR SDRAM: Dual rate synchronous dynamic RAM is similar to SDRAM, but has higher bandwidth, that is, faster.The maximum transfer rate to L2 cache is about 1064MB per second.(133 MHz DDR SDRAM).
RDRAM: Rambus dynamic random access memory is fundamentally different from the previous DRAM system.The RDRAM designed by Rambus Company uses Rambus in-line memory module (RIMM), which is similar to standard DIMMs in terms of overall dimensions and pin structure.The difference of RDRAM is that it adopts a special high-speed data bus design, called Rambus channel.The working frequency of RDRAM memory chip in parallel mode can reach 800 MHz (data rate 1600 megabytes).Due to the high operating speed, RDRAM generates much more heat than other types of chips.To dissipate excess heat, the Rambus chip is equipped with a heat sink that looks like a long, thin disk.Just as the DIMM has its small form factor version, the manufacturer has also designed a small form factor RIMM for notebook computers.
Credit card memory: The credit card memory is a patented independent DRAM memory module, which should be inserted into the special long slot of the laptop when used.
PCMCIA memory card: another independent DRAM memory module for notebook computers. This memory card does not have patent rights. As long as the system bus can match the memory card settings, it can be used for various notebook computers.
CMOS RAM: The term CMOS RAM refers to a small capacity memory used in computers and other devices to store information such as hard disk settings. For more information, see Basic Computer Knowledge.This type of memory requires a small battery to power it to maintain its contents.
VRAM: Video RAM, also known as Multi Port Dynamic Random Access Memory (MPDRAM), is dedicated to display adapters and 3D accelerators.The so-called "multi port" means that VRAM usually has two independent access ports instead of a single port, allowing CPU and graphics processor to access RAM at the same time.VRAM is located on the graphics card, and there are many kinds of VRAM, many of which have patent rights.The size of VRAM often determines the resolution and color depth of the display.VRAM can also be used to save some graphics specific information, such as 3D geometric data and quality maps.Real multi port VRAM is often expensive, so today's graphics cards use SGRAM (synchronous graphics RAM) as a substitute.The performance of the two video memories is almost the same, and SGRAM is cheaper.
