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

In this issue, the editor will bring you the relevant principles and classification. The article is rich in content and analyzes and narrates it from a professional point of view. I hope you can get something after reading this article.

There are many types of storage, including ROM (Read-onlymemory read-only), RAM (Random-accessmemory readable and writable), and a neglected CAM (Baidu can be used on its own).

Another method on the Internet classifies SRAM/DRAM/DDRAM as RAM and ROM/EEPROM/HDD/FLASH as ROM. In fact, this classification method is roughly based on the criteria of whether a power outage loses information, rather than a simple readonly.

The above methods are not used for classification below.

At present, most of the mainstream memories are RAM, and according to the principle of RAM, they are also divided into two types-volatile (Volatile) and non-volatile (Non-Volatile). The difference lies in whether to save data after power off. Volatile memories include SRAM, DRAM, SDRAM and DDR, which are mainly used for cache and memory strips, respectively. Non-volatile memory mainly consists of hard disk (magnetism, HardDisk Drive, HDD), Flash, optical disc (optical), used in our U disk, SD card and SSD hard disk. Optical discs and magnetic hard disks are not covered below.

01 、 ROM

ROM is the abbreviation of English Read-OnlyMemory. Translated into Chinese, it means "read-only memory". The computer term is "read-only memory". The contents of this kind of memory are written in advance by electronic process after people have made it. After that, you generally can not modify the contents of it, but can only read from it, but there are also erasable ROM, in which the data will not be dropped.

2. Non-volatile RAM2.1 principle

The standard physical structure of FlashMemory is called bit (cell), which is characterized by the fact that the Gate of the general MOS and the interval between the channel is the oxide insulation (gateoxide), while the FlashMemory has an extra layer of material between the Controlgate and the channel, which is called the floating gate (floatinggate). Thanks to this floating gate, FlashMemory can complete three basic modes of operation, namely, read (byte or word), write (byte or word), and erase (one or more memory spaces). Even without providing power to memory, data integrity can be preserved through this floating gate.

NorFlash

NORFlash takes a long time to erase, but it provides complete addressing and data bus and allows random access to any area of memory, which makes it very suitable to replace old-fashioned ROM chips. NORFlash can tolerate 10, 000 to a million write cycles, and it was the foundation of early removable flash storage media.

After the general introduction of the physical structure, the agreements of the various associations are left.

The full name of MMC is "MultiMediaCard", and MMC is a communication protocol that supports two modes, SPI and MMC. MMC mode is the standard default mode with all the features of MMC. The SPI mode is the second optional mode for MMC memory cards, which is a subset of the MMC protocol.

The SD card introduced below, emmc,UFS is the IC chip, in detail, NANDFlash is a storage medium, to read and write data on it, external control and circuit design are added, eMMC is the NANDflash+ master IC, and the external interface protocol is similar to SD and TF cards; for manufacturers, the circuit design is simplified and the cost is reduced.

SD card data transfer and physical specifications are developed from MMC and are about the same size as MMC. The length and width are the same as MMC, but a little thicker than MMC. In terms of compatibility, SD cards are downwards compatible with multimedia cards (MultiMedia Card). Therefore, the SD card also supports SPI interface access.

Emmc memory chip simplifies the design of memory, encapsulates NANDFlash chip and control chip with MCP technology, saves the area of circuit board consumed by zero components, and greatly improves the convenience of mobile phone manufacturers or computer manufacturers in designing new products. EMMC integrates FlashController in it, including protocol, erase-write equalization, bad block management, ECC check, power management, clock management, data access and other functions. Compared with directly connecting NANDFlash to host, eMMC shields the physical characteristics of NANDFlash, which can reduce the complexity of software on Host, let Host focus on upper-level business, and save special processing for NANDFlash. At the same time, eMMC is much better than NAND Flash in reading and writing performance by using Cache, MemoryArray and other technologies. On the other hand, the reading and writing speed of emmc is also faster than that of NANDFlash, and the reading and writing speed of emmc can be as high as 50MB to 100MB per second.

Students who follow the mobile phone circle often hear the word UFS. Let's talk about the relationship between UFS and emmc.

On computers, from HDD to SSD, from SATASSD to PCIeSSD, hard drives are getting faster and faster.

On mobile phones, from SD card to eMMC card to UFS card, the speed of memory card is also getting faster and faster.

3D NANDflash is not a stack of chips, but a direct stand-up of NAND. In this way, the number of transistors that exist per unit area is the number of stacks. So Moore's Law expects that the area of the transistor will be converted to an increase in the number of stacking layers. At present, it is said that the NANDflash with 64 or more layers has been put into mass production.

03. Volatile RAM

The two main types of volatile random access semiconductor memory are static random access memory (SRAM) and dynamic random access memory (DRAM). The commercial use of semiconductor RAM dates back to 1965, when IBM introduced SP95SRAM chips for their System/ 360Model 95 computers, while Toshiba used DRAM memory cells for its ToscalBC-1411 electronic calculators, both based on bipolar transistors. Commercial MOS memories based on MOS transistors were developed in the late 1960s and have been the basis of all commercial semiconductor memories ever since. The first commercial DRAMIC chip, Intel1103, was launched in October 1970. Synchronous dynamic random access memory (SDRAM) then made its debut with Samsung KM48SL2000 chips in 1992.

3.1 principle

Memory unit is the basic building block of computer memory. The memory cell is an electronic circuit that stores a bit of binary information, which must be set to store logic 1 (high voltage level) and reset to store logic 0 (low voltage level). Its value is maintained / stored until it is changed through the set / reset process. The value in the storage unit can be accessed by reading.

In SRAM, memory cell is a kind of flip-flop circuit, which is usually implemented by FET. This means that SRAM requires very low power consumption when not being accessed, but it is expensive and low storage density.

The second type, DRAM, is based on capacitors. Charging and discharging the capacitor can store "1" or "0" in the battery. However, the charge in the capacitor will slowly leak out and must be refreshed regularly. As a result of this refresh process, DRAM uses more power, but it can achieve higher storage density and lower unit cost than SRAM.

The above is the principle and classification shared by the editor. If you happen to have similar doubts, you might as well refer to the above analysis to understand. If you want to know more about it, you are welcome to follow the industry information channel.

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.