Home  Memory

Oct 12 2019

Classifications and Characteristics of Computer Memory in Type Principles


The memory is an entity that memorizes information, which is important for the computer to have the ability to store data and information, to automatically execute programs continuously, and to perform extensive information processing. The computer memory is composed of a plurality of storage units, and the computer reads and writes data from the memory in storage units, that is, one storage unit data is read at a time necessarily. In order to distinguish between different storage units, the system assigns each storage unit a unique number, which is called the address of the storage unit, that is, the storage address.

How does memory work in a computer system? This video talks about cells, information stored in those cells, addresses, sizes and how data gets stored in memory.



Classification Principles


Classification Standards

Memory Cards


Memory Capacity



Hybrid Memory




1) According to the Media of the Computer Memory

2) According to the Data Access Mode of the Memory

3) According to Its Role in Computer Systems

4) Note:

Main Memory

1) Memory Applied in SiP

2) Compact Flash Card

3) Smart Media Flash Card

4) Memory Stick Duo

5) Secure Digital Flash Card 

6) XD Flash Card

7) Multimedia Card

8) MicroDrive

9) USB-CF-Card

10) Digital Film

11) PC Card Converter

Computer Memory Grade

Ⅰ Classification Principles     

Different devices require different computer memories, and the most commonly used memory types in embedded systems fall into three categories.

a. RAM(random-access memory)

The RAM can read or write data at any address at any time.

The advantage of RAM is that it is easy to read/write and flexible to use.

The disadvantage of RAM is that it cannot save information for a long time, in other words, once power is lost, the stored data will be lost.

RAM is used for temporary storage or buffer storage of binary information.

The RAM is further divided into SRAM (static random-access memory) and DRAM (dynamic random-access memory).

SRAM saves data by transistor latching. It has a complicated process and high production cost. Its storage speed is faster, so it is generally used as cashe as a bridge between CPU and internal memory, such as the L1 Cache and L2 Cashe in the processor. Due to process characteristics, SRAM integration is not very high, so its storage capacity is generally small.

DRAM (dynamic random-access memory) saves data by capacitor charging. DRAM are more common than SRAM, because the electric memory used in computers is DRAM. With the development of technology, DRAM has developed into SDRAM (synchronous dynamic random access memory) and DDR SDRAM ( double rate synchronous dynamic random-access memory). SDRAM only represents one data on the rising edge of the clock, while DDR SDRAM can represent one data on both rising and falling edges.

DDR has been developed into DDR2, DDR3, and DDR4. On this basis, in order to meet the low power requirements of mobile devices, LPDDR (low power double data rate SDRAM) has been developed. The further development of LPDDR technology includes LPDDR2, LPDDR, and LPDDR4 respectively.

In addition, the most frequently used memory applications in mobile phones are LPDDR3 and LPDDR4, and the mainstream configuration is 3G or 4G capacity at present.


b. ROM(read-only memory)

ROM used to refer to read-only memory, which can only read data in it and cannot write data to it. Therefore, this kind of electric memory data is already written by the manufacturers, which can’t be modified again. The common application is the BIOS in the computer. With the development of technology, ROM can be written data in it, but the name has been retained.

The data stored in the ROM can be arbitrarily read. After the power is turned off, the data in the ROM remains unchanged, but the data cannot be written.

ROM is very useful in embedded systems and is often used to store system software (such as ROM BIOS), applications, and other code or data that does not change over time.

ROM memories can be classified based on development: mask ROM, programmable ROM (PROM), and erasable programmable ROM (EPROM).

Commonly used in ROM are EPROM and EEPROM.

EPROM is a ROM memory that has an erasable function and can be reprogrammed after erasing. Before writing, the data must be cleared by ultraviolet light to illuminate the transparent window on the IC. This type of chip is relatively easy to identify, and its package contains quartz glass window. The window of a programmed EPROM chip is generally covered with black self-adhesive paper to prevent exposure to ultraviolet light.

The EPROM can be erased multiple times and then written multiple times. However, it is necessary to erase under a specific environment of ultraviolet rays, so this memory is not easy to write.

EEPROM (electrically erasable programmable ROM), electrically erasable ROM, is now used more, because as long as there is electricity, the data can be erased, and then the data is rewritten, that is, the programming can be repeated repeatedly during use.

c. Hybrid Memory

The hybrid memory can be read and written at random, and the data in the device can be kept unchanged after the power off. And hybrid storage devices can be divided into three types:


EEPROM is an electrically erasable programmable memory device. Unlike EPROM, EEPROM uses electricity to remove data, rather than ultraviolet radiation.

The EEPROM allows the user to erase and rewrite the data multiple times in bytes, and can be directly carried out in the machine. It does not require special equipment, is convenient and flexible, and is often used for data and parameters that are frequently modified and require power failure protection requirements.


NVRAM is usually regarded as an SRAM with a backup battery. When the power is turned on, the NVRAM is just like any other SRAM, but when the power is turned off, the NVRAM gets enough power from the battery to keep the existing contents.

(3) Flash

Flash memory is an EEPROM that does not require a Vpp voltage signal. A sector of a byte can be erased in an instant (a very short time compared to a single clock cycle).

In other words, flash is a variant of EEPROM, the difference is, EEPROM can be deleted and rewritten at the byte level instead of the entire chip erase, but most flash chips require block erase. In addition, compared with EEPROM, flash has a larger storage capacity.

The advantage of Flash over EEPROM is that it can erase many bytes at the same time, saving the time of erasing to write data, but once a sector is erased, it must be written byte by byte, and its write time is so long.

FLASH is widely used. U disk, CF card, SM card, SD / MMC card, memory stick, XD card, MS card, TF card, etc. are all FLASH, except that, SSD solid state drive also belongs to FLASH.

cache memory

Flash is divided into Nor Flash and Nand Flash:

Intel first developed Nor Flash technology in 1988, completely changing the situation that was originally dominated by EPROM and EEPROM. In 1989, Toshiba released the Nand Flash structure, emphasizing the cost per bit reduction and having higher performance. And like a disk, you can easily upgrade through the interface.

Nor Flash is different from Nand Flash. Nor Flash is more like internal memory. It has independent address lines and data lines, but the price is more expensive and the capacity is smaller. Nand Flash is more like hard disk, address line and data line are the I/O line. All information is transmitted through a hard disk cable, and it has a lower cost and a much larger capacity than Nor Flash.

If flash is only used to store a small amount of code, then Nor Flash is more suitable. Nand Flash is the ideal solution for large data storage.

Therefore, Nor Flash type flash memory is more suitable for frequent random read and write occasions, usually used to store program code and run directly in flash memory. Nand Flash type flash memory is mainly used to store data. Our commonly used flash memory products, such as U disk and memory card both use Nand Flash.

Running the code on Nor Flash does not require any software support, but drivers are usually required for the same operation on Nand Flash.

At present, the memory capacity of the smart phone in the mobile phone is relatively large, the mainstream configuration has 32G~256G storage space at present, and the Nand Flash is usually used, in addition, the external expansion memory card of the mobile phone is also Nand Flash from before.

 computer memory

Ⅱ Specific Classification Standards

  • According to the media of the computer memory

1. Semiconductor memory

A memory composed of a semiconductor is called a semiconductor memory, and this kind of semiconductor memory is small in size, low in power, and short in access time. However, when the power is off, the stored data is also lost, which is a kind of volatile memory.

2. Magnetic memory

Magnetic memory is made of magnetic materials. In general, a layer of magnetic material is coated on metal or plastic to store data. The characteristic is non-volatile, and lower access speed. 

3. Disk storage

The hard disk is one of the main storage media of the computer, and is composed of one or more discs made of aluminum or glass, and the disk is covered with a ferromagnetic material. The hard disk is also the storage with the largest storage capacity.

The optical disc is a medium that uses optical information as a storage carrier and is used for storing data. It is a device for reading and writing by using the laser, and is a rapidly developed auxiliary storage, which can store various characters, sounds, graphics, multimedia digital information such as images and animations. Because the optical disc memory uses a laser to read on a magneto-optical material, which is characterized by non-volatility, good durability, and high recording density. And it is now used in computer systems as external storage.

The disk is divided into CD, DVD, Blu-ray disc, etc. The capacity of the CD is only about 700MB, while the DVD can reach 4.7GB, and the Blu-ray disc can reach 25GB. The difference in capacity between them is closely related to the wavelength of the associated laser beam. The shorter the wavelength, the greater the storage density and the larger the capacity of the disc.


  • According to the data access mode of the memory

1. RAM (random memory)

It is a memory that can be read and written. The content of any one of its storage units can be randomly accessed, and the access time is independent of the physical location. Our main memory is this kind of RAM.

2. ROM (Read Only Memory)

It is a type of original information that can only be written once. After writing, only the internal data can be read, but cannot be rewritten at will, that is, there is on new data to change the original information.

3. Serial access data memory

When reading and writing to a memory location of a memory, the address must be addressed sequentially according to the physical location of the memory unit. This memory is a serial access memory. When such a memory accesses data, it needs to be accessed in accordance with the location display of the memory location.


  • According to its role in computer systems

1. Main memory

It usually refers to the internal memory we call, it can directly exchange data with the CPU memory, featuring fast speed, small capacity and high price. The main memory is made of semiconductor, so it is a volatile memory.

2. Auxiliary memory

It usually refers to what we call external memory, used to store programs and data that are not currently used. It cannot directly exchange data with the CPU and needs to be loaded into main memory. Features slow speed, large capacity and low price. It is a non-volatile memory.

3. Buffer memory (cache)

It mainly used in two components with different speeds. Now it is basically used between CPU and main memory to play the role of caching.


——Note: main memory


The main memory is a memory that can be directly written by the CPU. It stores the programs that need to be executed and the data that needs to be processed. It can only temporarily store data and cannot store data for a long time.


1) Memory bank (MPS)

It consists of memory cells (each cell contains several storage components, each component can store a binary number) and each cell has a number, called the memory cell address. Usually a memory unit consists of 8 storage elements.

2) Address register (MAR)

It consists of several flip-flops, which are used to store the data access address, and the length of the address register matches the register capacity (ie, the capacity is 1K, and the length is  2^10=1K).

3) Address decoder and driver

4) Data register (MDR)

The data register consists of a number of flip-flops, which are used to store the data read in the memory cell, or temporarily store the data to be written to the memory cell from the data bus [width of the data memory should match the length of the storage unit].

electric memory

Ⅲ Memory Cards

There are many different memory cards applied in reality based on practicle requirements, the following are introducing them:

1) Memory applied in SiP

In SiP design, integrated memory is also generally required. Knowing more characteristics of  memory types, we can know which type of memory is integrated in the SiP project.

At present, it is common to integrate DDRx and Flash in SiP for program operation and data storage. Of course, SRAM can also be integrated into SiP to increase the speed of system operation.

The memory can be a card or a floppy disk, either active or fixed one for saving images. Here are some of memory types in real life.


2) Compact flash card

A compact flash card that plugs into a digital camera like a PC card. It can be used with an adapter (also known as a riser card) to adapt to a standard PC card reader or other PC card devices.

Part of the cf memory card is made of tempered glass and metal casing. The cf memory card uses a standard ata/ide interface and is equipped with a special pcm-cia adapter (transfer card). Laptop  users can use it directly on the pcmcia slot, making data transfer between digital cameras and computers conveniently.


3) Smart media flash card 

Smart media flash card, is a storage medium. SM card uses ssfdg/flash memory card, with characteristics of ultra-small, ultra-thin, ultra-light, low power consumption, and easy to upgrade. In addition, SM conversion card also has a pcmcia interface for easy data transfer.


4) Memory stick duo

Memory stick duo is about one-third the size and weight of ordinary memory sticks, and the current maximum storage capacity can reach 4g.


5) Secure digital flash card  

The storage speed is very fast, very small, and the appearance is the same as that of MMC. More digital cameras on the market use memory cards of this format.


6) XD flash card  

Fuji Film and OLYMPUS jointly launched the xD-Picture card, which is small in size and fast in transmission, but it is expensive.


7) Multimedia card

The appearance is exactly the same as SD, and it is often used in general.


8) MicroDrive

It is a relatively high-end storage product. The micro hard disk has the same appearance as the CF card and uses the same model interface.


9) USB-CF-Card

It is a digital camera storage medium produced by Lexar. It has the same appearance as a general cf card. It can be used in digital cameras such as cf cards, pda, mp3, etc., and can be connected directly to the computer system via the usb interface used as a mobile storage.

 USB stick

10) Digital film 

A storage medium for a digital camera, which is a digital storage medium of the same type as a Hitachi sm card, a Panasonic sd card, and a Sony memory stick.


11) PC card converter

A connector that can insert a cf card or sm card into it, and then insert the whole as a pc card into the pcmica socket of the computer. This is a general expansion interface commonly used in portable computers, and can also insert a pccica memory card, a pcmica hard disk, and pcmica modem into it


Ⅳ Memory Capacity

(1) The level of memory is divided according to its three indicators: speed, capacity, and price per bit:

Register => Cache => Main Memory => Disk => CD

The higher the upper memory, the smaller the capacity, the faster the speed, the higher the price per bit, and the larger the memory capacity of the lower layer, the slower the speed, the lower the price per bit.

The register is a memory in the CPU. The CPU actually calculates and controls the number in the register. It is the fastest and the most expensive. In addition, the cache is also set to the CPU.

(2) Cache and main memory are mainly to solve the problem that the CPU does not match the main memory speed, because the CPU speed is faster than the main memory, and the cache is fast than the main memory. As long as the data to be used by the CPU is transferred to the cache, the CPU can directly obtain data from the cache to increase the access speed of the data and reduce the load on the CPU, and the main memory and cache data transfer is done by the hardware itself.

(3) Main memory and auxiliary memory are mainly used to solve the capacity problem of the storage system. The auxiliary memory is lower than the main memory and cannot be accessed by the CPU, but its capacity is large. When the CPU needs to run the program, it will be auxiliary. The data is transferred to the main memory, and the CPU is coming to access it. The data transfer between the main memory and the auxiliary memory is performed by the hardware and the operating system.

Related Articles

  • Key Characteristics of Main Computer Mem...
    IntroductionIn computing system, Memory refers to the computer hardware integrated circuits that store information for immediate use in a computer. Like many things, computer memory comes in many diff...
  • What Are the New Memory Technologies and...
    Warm hints: This article contains about 5000 words and reading time is about 20 mins.IntroductionThe rapid development of information technology requires storage technology to provide faster, larger c...
  • Ferroelectric Random Access Memory (FeRA...
    Ⅰ. Introduction With the improvement of computer technology, the demand for non-volatile memory is increasing, their read and write speed requirements are getting faster and faster, an...
  • The Comparing of the NorFlash, NandFlash...
    This article is mainly about the relative introduction of the NorFlash, NandFlash and EMMC, and pays attention to the detailed differences of the three.NorFlashNorFlash is one of the two main technolo...
  • What is Computer Memory? Types of Comput...
    Ⅰ IntroductionThis video going to trace the history of these storage technologies from punch cards, delay line memory, core memory, magnetic tape, and magnetic drums, to floppy disks, hard disk drives...
  • New Type of Non-volatile Computer Memory...
    Ⅰ Introduction In recent years, with the rapid development of the internet of things (IOT) and electronic technologies, embedded devices such as mobile phones, smart watches, and sports bra...


    • Eille Wan on 2019-10-17 17:44:14

    After read your note, I add that DRAM uses the charge on the gate capacitance of the MOS transistor to store data. Once the power off, the information is lost, in addition, the gate will leak, so a refreshing mechanism is needed to recharge the gate capacitance at regular intervals. And each time the data is read out, it needs to be replenished. This is also called dynamic refresh, so it is called dynamic random access memory.

    • Maxu on 2019-10-28 13:48:12

    Hello, can you briefly say what the bit extension is?

      • author on 2019-10-28 17:16:37


      Bit expansion refers to increasing the memory word length, because the number of data lines of the memory chip is not necessarily equal to the number of data lines of the CPU. In this case, the memory chip needs to be bit-expanded so that the number of data lines and the number of bits of the CPU data line can be consistent.

    • Teaw on 2019-11-13 13:58:36

    There are relatively few MTDs required in NOR devices, and many vendors offer more advanced software for NOR devices, including M-System's TrueFFS driver, which is used by Wind River System, Microsoft, QNX Software System, Symbian, and Intel. In addition, the driver is also used to simulate DiskOnChip products and manage NAND flash, including error correction, bad block processing and wear leveling.

Leave a Reply

Your email address will not be published.