Shulou(Shulou.com)11/24 Report--

Earlier, I gave you a detailed introduction to the development history of HDD hard drives, floppy disks and CDs (links).

You should have noticed that in our daily life, there are far more than the above three storage media.

The U disk, TF card, SD card that we often use, as well as the DDR memory and SSD hard disk used on computers, all belong to another storage technology.

This technology, we call it "semiconductor storage".

Today, Xiao Zaojun will focus on telling you about this knowledge.

The classification of █ semiconductor storage modern storage technology, generally speaking, is divided into three parts, namely, magnetic storage, optical storage and semiconductor storage.

Semiconductor memory, in short, is a memory that uses "semiconductor integrated circuits" as the storage medium.

If you take apart your U disk or SSD hard drive, you will find that it is full of PCB circuit boards, as well as a variety of chips and components. Among them, there is a kind of chip, which is specialized in storing data, sometimes called "memory chip".

Construction of ▲ SSD hard disk

Semiconductor memory is lighter, smaller, and faster to read and write than traditional disks, such as HDD hard drives. Of course, the price is also more expensive.

In recent years, the whole society pays close attention to the chip semiconductor industry. However, people are mainly concerned about computing chips such as CPU, GPU, mobile phone SoC and so on.

As everyone knows, semiconductor memory is also one of the core pillars of the entire semiconductor industry. In 2021, the global market for semiconductor memory is US $153.8 billion, accounting for 33% of the total IC market, or 1/3.

The proportion of memory in the major categories of semiconductors worldwide declined in ▲ 2022, but still 26%.

Semiconductor memory is also a large category, which can be further divided into volatile (VM) memory and non-volatile (NVM) memory.

As the name implies, after the circuit is powered off, the volatile memory cannot retain the data, and the non-volatile memory can retain the data.

This is actually easier to understand. Children's shoes who have learned the basic knowledge of computer should remember that storage is divided into memory and external memory.

Memory used to be called running memory (operational memory). After the computer is powered on, it works with CPU and so on. After the power is off, the data is gone and belongs to volatile (VM) memory.

The external memory, that is, the hard disk, stores a large number of data files. When the computer shuts down, as long as you perform the save (write) operation, the data will continue to exist, belonging to non-volatile (NVM) memory.

Please note: now many materials also divide semiconductor memory into random access memory (RAM) and read-only memory (ROM). You should be familiar with it, right?

ROM read-only memory: easy to understand, can be read, can not be written.

RAM random access memory: refers to that it can "randomly read or write data from any memory cell", as opposed to the traditional magnetic storage must be "sequential access" (Sequential Access).

Some people think that volatile memory is RAM and non-volatile memory is ROM. As a matter of fact, this is not rigorous, and the reasons will be explained later.

█ volatile memory (VM) has not changed much in the past few decades, and it is mainly divided into DRAM (dynamic random access memory, Dynamic RAM) and SRAM (static random access memory, Static RAM).

DRAMDRAM consists of many repetitive Bit Cell, each of which consists of a capacitor and a transistor (also known as 1T1C structure). The amount of charge stored in a capacitor, used to represent "0" and "1". The transistor is used to control the charge and discharge of the capacitor.

▲ Picture Source: Lam Research

Because the capacitor will have the phenomenon of leakage. Therefore, periodic "dynamic" charging must be carried out before the data is changed or the power is cut off to maintain the potential. Otherwise, data will be lost.

Therefore, DRAM is called "dynamic" random access memory.

DRAM has always been the mainstream solution of computer and mobile phone memory. The memory bar of the computer (DDR), the video memory of the graphics card (GDDR), and the running memory of the mobile phone (LPDDR) are all a kind of DRAM. (DDR basically refers to DDR SDRAM, double-rate synchronous dynamic random access memory. )

It is worth mentioning that on the video memory side, in addition to GDDR, there is a new type of video memory called HBM (High Bandwidth Memory). It is made up of a lot of DDR chips stacked and encapsulated with GPU (no apparent memory particles can be seen).

SRAMSRAM, you may be strangers. In fact, it is the technology we use for CPU caching.

The architecture of SRAM is much more complex than DRAM.

The basic unit of the SRAM consists of at least six transistors: four field effect transistors (M1, M2, M3, M4) form two cross-coupled inverters, and two field effect transistors (M5, M6) are used to control the bit line (Bit Line) for reading and writing. Through these field effect transistors, a latch (flip-flop) is formed, and the binary numbers 0 and 1 are locked when powered on.

Therefore, SRAM is called "static random access memory".

▲ SRAM storage unit

SRAM does not need regular refresh and fast response, but it has the advantages of high power consumption, low integration and high price.

Therefore, it is mainly used for the primary cache and secondary cache of CPU. In addition, it will also be used in FPGA. Its market share has always been relatively low and its sense of existence is relatively weak.

█ non-volatile memory (NVM) next, take a look at the non-volatile memory products.

There are more technical routes for non-volatile memory products. The earliest one is the ROM mentioned above.

The oldest ROM, the "real" ROM-- is completely read-only. By the time it comes out of the factory, the storage content is dead and cannot be modified.

This kind of ROM is very inflexible, and if something is wrong, it can't be corrected and can only be discarded.

Mask model read-only memory (MASK ROM) is the representative of this kind of ROM. To put it bluntly, it is to directly use the mask process to "engrave" the information into the memory, so that the user can not change it, which is suitable for early mass production.

Later, experts invented PROM (Programmable ROM, programmable ROM). This kind of ROM can only be programmed once. When leaving the factory, all storage units are 1. Through special equipment, the fuse can be fused in the way of current or light (ultraviolet), and the effect of rewriting data can be achieved.

PROM is a little more flexible than ROM, but it's not enough. It is best to be able to modify the data, so some experts invented EPROM (Erasable Programmable, erasable programmable ROM).

The way of erasing can be light or electricity. Electricity is more convenient. Erasers using electricity are called EEPROM (electrically erasable programmable EEPROM).

EEPROM is modified in Byte as the minimum unit. In other words, you can write 0 or 1 to each bit, just press "bit" to read and write. You don't have to erase all the contents before writing. Its erasure operation, also in "bit" units, is still too slow.

In the 1980s, Fujio Masuoka, a technical expert at Toshiba in Japan, invented a new kind of memory that can be erased quickly, that is, Flash (flash memory).

▲ Fujio Masuoka

Flash means "quickly" in English.

Limited to space, we will specifically introduce the specific principles of FLASH next time. All we need to know is that Flash storage is erased in "blocks".

Common block sizes are 128KB and 256KB. 1KB is 1024 bit, which is several orders of magnitude faster than EEPROM erasing by bit.

At present, there are only two mainstream representative products of FLASH: NOR Flash and NAND Flash.

NOR FlashNOR Flash is a code-based flash memory chip, its main feature is on-chip execution (XIP,Execute In Place), that is, the application program no longer needs to read the code into the system RAM, but can run directly in the Flash flash memory.

Therefore, NOR Flash is suitable for storing code and part of data, with high reliability and fast reading speed, and has advantages in performance and cost in low and medium capacity applications.

However, NOR Flash is slow to write and erase, and is twice the size of NAND Flash, so its use is limited and its market share is relatively low.

In the early days, NOR Flash was still used in high-end phones, but later, when smartphones began to introduce eMMC, even this market was squeezed out.

In recent years, the application of NOR Flash has rebounded and the market has warmed up. Low-power Bluetooth module, TWS headset, mobile phone touch and fingerprint, wearable devices, automotive electronics and industrial control and other fields, more use of NOR Flash.

Compared with NAND Flash, NAND Flash has a much larger market share.

NAND Flash is a kind of data flash memory chip, which can realize mass storage.

It reads and writes data in pages and erases data in blocks, so although its writing and erasing speed is about 3-4 orders of magnitude slower than DRAM, it is also 3 orders of magnitude faster than traditional mechanical hard drives, and is widely used in eMMC / EMCP, U disk, SSD and other markets.

EMMC was mentioned earlier. A few years ago, the word was quite popular.

▲ eMMC

EMMC is embedded multimedia card (embedded Multi Media Card). It encapsulates MMC (multimedia card) interface, NAND and main controller in a small BGA chip, mainly to solve the problems of NAND brand compatibility, and to facilitate manufacturers to launch new products quickly and simply.

On the other hand, eMCP encapsulates eMMC and LPDDR as a whole, which further reduces the module size and simplifies the circuit connection design.

In 2011, the UFS (Universal Flash Storage, Universal Flash Storage) 1.0 standard was born. Later, UFS gradually replaced eMMC as the mainstream storage solution for smartphones. Of course, UFS is also based on NAND FLASH.

The standard configuration of mainstream mobile phones in ▲ these years

SSD, you should be familiar with it. It basically uses NAND chip, and it is developing very rapidly.

Internal structure of ▲ SSD

According to the difference of internal electronic cell density, NAND can be divided into SLC (single-layer memory cell), MLC (double-layer memory cell), TLC (three-layer memory cell and QLC (four-layer memory cell). The data stored in each memory cell is 1-bit, 2-bit, 3-bit and 4-bit respectively.

From SLC to QLC, the storage density increases gradually, and the unit bit cost decreases accordingly. But in contrast, performance, power consumption, reliability and P / E cycles (the number of erase and write cycles, that is, life) will decline.

In recent years, there has been a great controversy around SLC / MLC / TLC / QLC in the DIY installation circle. At first, netizens thought that the life of SSD hard drives would shrink. Later, it was found that the shrinkage was not so serious, and the life span was still sufficient. So, I slowly accepted it.

The early NAND was all 2D NAND. After the process enters 16nm, the cost of 2D NAND increases sharply, and the difficulty and cost of planar miniaturization process are difficult to bear. As a result, 3D NAND appeared.

▲ Picture Source: electronics-lab

To put it simply, from bungalows to buildings, three-dimensional stacking is used to increase memory capacity and reduce the process pressure of 2D NAND.

In 2012, Samsung launched the first generation of 3D NAND flash memory chips. Later, with the continuous development of 3D NAND technology, the number of stacking layers is increasing, and the capacity is also increasing.

█ New memory (non-volatile) in 2021, the American IBM put forward the concept of "SCM, Storage-Class Memory". IBM believes that SCM can replace traditional hard drives and complement DRAM.

Behind SCM is actually the industry's exploration of a new type of memory (media).

According to the consensus of the industry, the new memory can combine the high-speed access of DRAM memory, and the characteristic of NAND flash to retain data after power off, break the boundary between memory and flash memory, and combine them into one to achieve lower power consumption, longer life and faster speed.

At present, there are mainly several new types of memory: phase change memory (PCM), resistive memory (ReRAM / RRAM), ferroelectric memory (FeRAM / FRAM), magnetic memory (MRAM, the second generation is STT-RAM), carbon nanotube memory.

█ conclusion to sum up, Xiaozao Jun drew a complete classification diagram of semiconductor storage:

In the picture above, there are many types of memory. But as I said earlier, you can focus on DRAM, NAND Flash and NOR Flash. Because, in the current market, these three kinds of memory account for more than 96% of the market share.

In fact, all memories, based on their own characteristics, will find their own position in the market and give full play to their value.

In general, the more powerful the memory, the more expensive it will be and the closer it will be to the computing chip (CPU / GPU, etc.). The memory with weak performance can bear some requirements of low storage delay and insensitive writing speed, and reduce the cost.

Typical memory hierarchy in ▲ computer system

Photo source: fruit shell hard technology

In fact, the evolution of semiconductor storage technology has always benefited from Moore's Law, which continues to improve performance while reducing costs. In the future, with the gradual failure of Moore's Law, where will semiconductor storage technology go, and new storage media can rise? Let's wait and see.

References:

1. "data Survival White Paper", Huawei, Roland Berger

2. White Paper on China's Survival strength, 2022 arithmetic Conference

3. "computer Storage History", China Storage Network

4. "A brief History of hard disk Development", the I World of SunnyZhang

5. "the Development of Storage Technology", Xie Changsheng

6. "History of Storage Media", bilibili, cold attempt

7. Research report on next Generation data Storage Technology, Institute of Information and Communication

8. Memory Chip Industry Research report, Guoxin Securities

9. "domestic storage awaits a revolution", Fu Bin, fruit shell

10. "there is nothing more comprehensive about semiconductor storage than this one."

11. "Science and Technology Chapter 035-Flash memory of Semiconductor Storage", Wumijin, Zhihu

12. Wikipedia related entries.

This article comes from the official account of Wechat: fresh Jujube classroom (ID:xzclasscom), author: Xiaozaojun

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.