Start by making a simple science popularization. Changjiang Storage makes something called 3D NAND FLASH, and is a producer of high-performance SSD storage particles. There is also a company called Hefei Changxin in China, and what he makes is called Dram, which is the storage particles on the memory module. Taking the most commonly used mobile phones as an example, most Android machines now come standard with 8+128. This 8GB is called memory, and the big number 128 is called hard disk. All the programs in the phone run in memory, and the contents of the programs are stored on the hard disk.
So back to the point. According to Trendforce's calculation of sales data, in 2021, the global market share of Yangtze River storage will reach 3.4%. It is also expected that the market share of Yangtze River storage may reach 7% in 2022. At this time, there will definitely be a small partner who said, this is less than 5% na, or Others na, what is there to blow? Of course, it should be blown, because before 2020, the market share of Yangtze River storage is still 0%.
What is this concept?
Yangtze River Storage was established in 2016.
In 2017, the 32-layer MLC was successfully developed;
In August 2018, Xtacking technology was released to the public;
In 2019, the 64-layer TLC was successfully developed;
In April 2020, the company announced that the 128-layer TLC and QLC technology were successfully developed.
In October 2020, the 64-layer TLC NAND titanium pc005 was listed, representing the yield of the 64-layer TLC process of Yangtze River Storage entering the mass production level;
In September 2021, Asgard was launched with yangtze river storage 128-layer TLC NAND;
On December 29, 2021, Changcun released the latest series of its own brand, TIPro7000, which is also equipped with 128 layers of TLC particles. The two products represent the yield of the 128-layer TLC process of Yangtze River Storage officially entering the mass production level.
According to Techinsights, Yangtze River Storage's 128-layer process achieves new industry-leading standards in capacity, bit density, and I/O speed. Samsung Electronics CEO Jin Qinan said at the IEDM that the technology similar to Xtacking was originally ready to be launched in 2025, but we did not expect that Changcun took the lead in mass production. In other words, Yangtze River Storage spent 5 years to technically catch up with the traditional storage factories such as Samsung Hynix Magnesium, and even slightly with the meaning of overtaking in a little corner.
Samsung's xtacking-like technology won't be available until 2025
Let's go back to the topic of market share, what does it mean for the Yangtze River storage to come out? Does that mean Samsung won't be in a ceasefire? Obviously, this is a joke.
To answer this question, we must first understand a fact. In 2020, China spent more than 2.4 trillion yuan to import chips. Among them, about 700 billion yuan is used to import memory chips. Of the memory chips, about 130 billion yuan belongs to NAND FLASH. Before 2020, the 130 billion yuan was divided up by foreign companies such as Samsung, Micron, Hynix, Toshiba, and Intel. According to the instructions of the 14th Five-Year Plan, the future of China will be a digital China. The most important resource in the digital age is data, and data must have a place to store. As a result, many institutions predict that by 2026, the market space of China's NAND FLASH will exceed 300 billion.
■ In 2020, China's imported chips spent 2,420.7 billion yuan
The monthly production capacity of Changcun is now 100,000 pieces, and the second phase is planned to be 200,000 pieces, and the total production capacity is expected to reach 300,000 pieces in 2024. According to the current market share of 3.4%, by 2026, the market share of Yangtze River storage is likely to be close to 30%, or even higher. Then the long-term revenue until 2026 may exceed the level of 100 billion. What's more, the long-term breakthrough robbed Samsung of their jobs. You must know that most of the profits of Samsung Hynix come from memory chips. In the face of the two brothers of Changcun Changxin, the big dog of Samsung may not be able to withstand it.
I know that there will be many friends who are worried about the long-term future. After all, it only took 5 years to catch up, is this Xtacking technology fidelity? What if I get sanctioned by the United States?
First of all, this Xtacking technology must be fidelity. I open a storage company, can I still sell you fake technology? To discern whether the Yangtze River Storage Xtacking technology is 100% autonomous and controllable, it is necessary to start from another company, his name is Wuhan Xinxin. Students who understand Yangtze River Storage must be familiar with this company, because Yangtze River Storage was established on the basis of Wuhan Xinxin.
Founded in 2006 with a 10 billion yuan investment from the Wuhan Municipal Government, Wuhan Xinxin was the only wafer foundry in the central region at that time. At first, the plan was to make Dram, and then he went around and went to Spansion (acquired by Cypress in 2009) oem NAND Flash. In fact, it is a good thing, but it is bad luck, and it has suffered a financial crisis in less than two years. Spansion was no longer working, and Wuhan Xinxin was also uncomfortable. At this time, TSMC, Micron and other companies have looked at the fat meat of the new core. The Wuhan government thought about it again and again and still did not sell it to them, and finally chose SMIC, which has always had a commission agreement, and retained the flame of the new core. I have to say that in that era, it was really not easy to adhere to the idea of autonomy. Give Wuhan a big thumbs up!
In August 2013, Spansion decided to license MirrorBit, the underlying technology of 3DNAND, to Wuhan Xinxin. As a result, Xinxin has the basic ability to develop 3D NAND. The problem is that Spansion himself has never made 3DNAND. The situation faced by the new core is equivalent to building a house and digging a pit, but the house above is completely unknown about what to build and how to build it.
So who stepped forward to rescue Xinxin at this time? Is it IBM? No, IBM does have patent licenses with Xinxin. However, this authorization mainly focuses on low-power 65-nanometer RF and 45-nanometer low-power technologies. What really makes the new core technology leap forward is actually the Chinese Academy of Sciences.
On September 9, 2013, Wuhan Xinxin signed an IP joint authorization agreement with the Institute of Microelectronics of the Chinese Academy of Sciences, Tsinghua University, Fudan University, and Shanghai Institute of Microsystems. The agreement stipulates that Wuhan Xinxin will obtain the overall use rights of the four institutions' nearly 1,500 joint patent pools containing integrated circuit innovation technologies below 20 nanometers.
■ Wuhan Xinxin reached an IP authorization agreement with the Institute of Microelectronics of the Chinese Academy of Sciences
The history of the Microelectronics Institute's research on 3DNAND can be traced back to 2010, when they had just applied for the first "three-dimensional memory manufacturing" patent in China. The patent inventors are Huo Zongliang and Liu Ming. Huo Zongliang should now be the vice president of Yangtze River Storage, while Liu Ming became an academician of the Chinese Academy of Sciences in 2015.
■ China's earliest "three-dimensional memory manufacturing" patent
After traversing the past 15 years in China to apply for "three-dimensional memory" related patents, I found that the traditional storage manufacturers such as Samsung, Hynix, and Micron applied for three-dimensional memory-related patents in the same period. If you are a competitor to The New Core, will you use the patent as a weapon or give it to your opponent? At the same time, in a 2013 paper, I found that the Institute of Microelectronics began to study wafer bonding system-in-packages for three-dimensional memory around 2012.
■A review of embedded flash memory technology published by the Institute of Microelectronics in 2013
Therefore, Xtacking must be China's independent original technology, and it must be completely autonomous and controllable. But there will still be some pressure in the face of U.S. sanctions for a long time. Although the storage unit of NAND FLash does not have high requirements for the process, the logic circuit generally needs to be made with a more advanced process of about 20nm. This gives the possibility of U.S. sanctions surviving. However, I personally think that this is relatively unlikely. Although the domestic can not do 5nm, hard concave 20nm is still no problem. In addition, the speed of our entire integrated circuit progress recently is still relatively fast, so the long-term benefits of US sanctions will be very small, and even make american equipment companies lose blood. After all, The long-term existence is at the time of expansion and requires a large number of new equipment.
So, the technology is controllable autonomously, the equipment will be supplied normally, and the long-term future will be great!
Write at the end
Accidentally said a lot, even a little incoherently. But in the final analysis, it is because my wife likes the Yangtze River storage too much. Since the establishment of Wuhan Xinxin in 2006, Yangtze River Storage has spent 15 years to achieve technical juxtaposition with the world's top semiconductor storage companies, and even slightly catch up.
This is the first Chinese company to catch up with the technology of the top foreign companies in the semiconductor field since the founding of New China in 1949.
Yangtze River Storage is worth leaving a mark on the meritorious book of China's semiconductor industry, and it is worthy of praise by everyone who pays attention to the development of China's semiconductor industry.
I have bought titanium pc005 and am ready to come back with a tipro7000. Friends, support domestic production, starting from me.
Live on, come on.