China's aggressive push to develop a domestic semiconductor industry has largely been successful. The country now has fairly advanced fabs that can produce logic chips using 7nm-class process technologies as well as world-class 3D NAND and DRAM memory devices. However, there are numerous high-profile failures due to missed investments, technical shortcomings, and unsustainable business plans. This has resulted in numerous empty fab shells — zombie fabs — around the country, according to DigiTimes.

As of early 2024, China had 44 wafer semiconductor production facilities, including 25 300-mm fabs, five 200-mm wafers, four 150-mm wafers, and seven inactive ones, according to TrendForce. At the time, 32 additional semiconductor fabrication plans were being constructed in the country as part of the Made in China 2025 initiative, including 24 300-mm fabs and nine 200-mm fabs. Companies like SMIC, HuaHong, Nexchip, CXMT, and Silan planned to start production at 10 new fabs, including nine 300-mm fabs and one 200-mm facility by the end of 2024.

Plenty of zombie fabs

However, while China continues to lead in terms of new fabs coming online, the country also leads in terms of fab shells that never got equipped or put to work, thus becoming zombie fabs. Over the past several years, around a dozen high-profile fab projects, which cost investors between $50 billion and $100 billion, went bust.

Many Chinese semiconductor fab projects failed due to a lack of technical expertise amid overambitious goals: some startups aimed at advanced nodes like 14nm and 7nm without having experienced R&D teams or access to necessary wafer fab equipment. These efforts were often heavily reliant on provincial government funding, with little oversight or industry knowledge, which lead to collapse when finances dried up or scandals emerged. Some fab ventures were plagued by fraud or mismanagement, with executives vanishing or being arrested, sometimes with local officials involved.

To add to problems, U.S. export restrictions since 2019 blocked access of Chinese entities to critical chipmaking equipment required to make chips at 10nm-class nodes and below, effectively halting progress on advanced fabs. In addition, worsening U.S.-China tensions and global market shifts further undercut the viability of many of these projects.

So, let's go over some of China's most ambitious fab projects, many of which have fallen into oblivion, or have become a dreaded zombie fab.

Failures to learn from

Leading chipmakers, such as Intel, TSMC, Samsung, or SMIC have spent decades developing their production technologies and gain experience in chips on their leading-edge nodes. But Chinese chipmakers Wuhan Hongxin Semiconductor Manufacturing Co. (HSMC) and Quanxin Integrated Circuit Manufacturing (QXIC) attempted to take a shortcut and jump straight to 14nm and, eventually, to 7nm-class nodes by hiring executives and hundreds of engineers from TSMC in 2017 – 2019.

HSMC was founded in late 2017 with a plan to build 14nm and 7nm-capable logic fabs in Wuhan with an initial investment of around $19 billion. However, a land-use dispute halted construction in November 2019, and by mid‑2020, it suffered severe underfunding of billions of dollars. By March 2021, the local government seized the project, fired all employees, and confirmed no chip production had ever occurred.

QXIC similarly aimed at 14nm-class production when it was founded in 2019 as a sister venture to HSMC in Jinan, Shandong. The company was born after issues with HSMC occurred. Despite government backing, the project never progressed beyond hype: there were no equipment orders, no factory construction, and by 2021, operations were suspended. Interestingly, Cao Shan, who served as chief executive of QXIC, was also a former board member of HSMC.

Perhaps, the most notorious China fab venture failure — the first of many — is GlobalFoundries' project in Chengdu. GlobalFoundries unveiled plans in May 2017 to build an advanced fabs in Chengdu in two phases: Phase 1 for 130nm/180nm-class nodes and Phase 2 for 22FDX FD-SOI node. The company committed to invest $10 billion in the project, with about a billion invested in the shell alone.

Financial troubles forced GlobalFoundries to abandon the project in 2018 (the same year it ceased to develop leading-edge process technologies) and refocus to specialty production technologies. By early 2019, the site was cleared of equipment and personnel, and notices were issued in May 2020 to formally suspend operations.

The site and unfinished building remained uninhabited for five years before Shanghai Huali Microelectronics Corp. (HLMC), controlled by the Hua Hong Group, announced it would take over the dormant site in mid-2023. HLMC is one of a few Chinese companies that intend to develop a sub-10nm-class fabrication process. However, it is unclear whether the Chengdu fab will be used as its flagship facility. GlobalFoundries’ Chengdu project serves as a rare example of a recovery among China’s stalled semiconductor projects. A rare exception to the numerous failures that China has encountered thus far.

Dehuai Semiconductor, an analog and mixed signal IDM startup, was not so lucky. The company was launched in 2019 with the help of local authorities. Dehuai did not present a clear roadmap, and made false claims about how its project was proceeding. By mid-2021, local anti-corruption authorities arrested key executives after investigations revealed that no fab had been built: only initial site preparation had begun. The project was one of the most notorious examples of fraud and mismanagement among China’s failed semiconductor ventures.

The fate of Fujian Jinhua Integrated Circuit (JHICC) is a bit different. Formally, this is not a failed project, but it is not a living one either. JHICC was launched with an ambition to build China's first large-scale DRAM fab in 2016. The company magically began trial production about two years after its inception, but it was soon discovered that it had stolen process technology from Micron using the help of UMC. Eventually, the U.S. Commerce Department put Fujian Jinhua into its Entity List, cutting its access to any American technology. This essentially stops the development of new process technologies and bans the procurement of any advanced tools. As a result, while JHICC has formally survived and exists on paper, it is a spectre of its former ambitions.

Another memory project that has failed in China is Jiangsu Advanced Memory Semiconductor (AMS). The company was established in 2016 with the plan to lead China's efforts in phase-change memory (PCM) technology. The company aimed to produce 100,000 300-mm wafers annually and attracted an initial investment of approximately $1.8 billion. Despite developing its first in-house PCM chips by 2019, AMS ran into financial trouble by 2020 and could no longer pay for equipment or employee salaries. It entered bankruptcy proceedings in 2023, and while a rescue plan by Huaxin Jiechuang was approved in 2024, the deal collapsed in 2025 due to unmet funding commitments.

Producing commodity types of memory is a challenging business. Tsinghua Unigroup was instrumental in developing Yangtze Memory Technology Co. and making it a world-class maker of 3D NAND. However, subsequent 3D NAND and DRAM projects were scrapped in 2022, after the company faced financial difficulties one year prior.

Tsinghua Unigroup’s second 3D NAND project aimed to mirror YMTC's model. But, at the time, even YMTC itself was still far from challenging multinational 3D NAND makers. So, the logic of building another costly fab (potentially reaching $24 billion) and possibly developing a new 3D NAND process technology was questionable.

For its DRAM efforts, Tsinghua brought in former Elpida CEO Yukio Sakamoto, who had experience competing with Samsung and Micron. However, he left in 2021 as Tsinghua approached bankruptcy, before he could contribute. Given the years and billions needed to develop DRAM technology — compounded by tool supply risks — Tsinghua scrapped its DRAM ambitions.

Logic and memory require rather sophisticated process technologies, and fabs that cost billions. By contrast, CMOS image sensors (CIS) are produced using fairly basic production nodes and on relatively inexpensive (yet very large) fabs. Nonetheless, this did not stop Jiangsu Zhongjing Aerospace, Huaian Imaging Device Manufacturer (HiDM), and Tacoma Semiconductor from failing. None of their fabs have been completed, and none of their process technologies have been developed.

China's failures could come back to haunt future ambitions

Typically, the more charges a NAND flash cell can hold, the less durable it is in terms of program/erase cycles. But 3D NAND material innovations, advancements of NAND controllers, and error correction algorithms can significantly increase the number of P/E cycle a NAND flash cell can sustain. This is what happened with Yangtze Memory Technologies Co.'s X3-6070 3D QLC device that boasts the endurance of 3D TLC ICs, ITHome reports. 

YMTC's X3-6070 3D QLC NAND device belongs to the company's fourth Generation products and features 128 active layers as well as Xtacking 3.0 architecture with a 2400 MT/s interface. While 128 active layers do not seem like a record by today's standards, one of the key part about this 3D QLC NAND device is that the manufacturer claims rather significant endurance of 4,000 program/erase cycles for this IC. The fast interface supported by these devices makes them suitable for the best SSDs featuring a PCIe 4.0 or PCIe 5.0 interface. 

Back in the day, it was believed that 3D TLC NAND could sustain 1,000 to 3,000 P/E cycles, but advancements of materials, controllers, and ECC methods increased that number to 4,000 P/E cycles and beyond. 3D QLC NAND was originally thought to be capable of 100 to 1,000 P/E cycles, but all makers of memory increased that number, as well. YMTC is perhaps the only company to advertise that its 3D QLC NAND is as durable as 3D TLC NAND, other makers of flash memory are surely also making progress in this direction.

Without any doubt, 4,000 P/E cycles is a great achievement for YMTC's X3-6070 3D QLC NAND memory devices. It remains to be seen how competitive these flash memory ICs are considering the fact that they have 128 active layers, whereas YMTC's competitors offer QLC 3D NAND devices with 176 or more active layers. 

YMTC currently uses its X3-6070 3D QLC NAND memory for its PC41Q consumer-grade solid-state drive. This SSD boasts a sequential read/write speed of 5500 MB/s and a data retention period of one year at 30 degrees Celsius, which is on par with TLC solid-state drives. With endurance of 4,000 P/E cycles, Yangtze Memory is looking forward to use 3D QLC NAND flash for enterprise-grade storage devices.

China's champion 3D NAND producer Yangtze Memory was recently accused by the Pentagon of being a 'military company', potentially posing risks to U.S. national security. In a strongly worded statement passed to Reuters, YMTC said that it does not supply its memory for military use.

"We have not supplied, or been directed by any entity, to supply our technology for military use," a statement by YMTC published by Reuters reads.  

YMTC is among the global 3D NAND technology leaders with its unique Xtacking technology that enables it to offer 3D NAND with range-topping speed data transfer speeds (potentially to enable the best SSDs with a PCIe 5.0 x4 interface). Therefore it is certainly not in YMTC's interests to upset the U.S. government by supplying memory to the People's Republic's Liberation Army. 

Meanwhile, being blacklisted by the U.S. Department of Commerce, and not exactly welcome by customers in Europe or the U.S., YMTC must be prepared to sell its 3D NAND memory exclusively to buyers in China. Some of those buyers are makers of SSDs, PCs, and smartphones and this is where Yangtze Memory can at least partly control the destiny of its 3D NAND devices.  

But YMTC produces loads of memory, so it has to sell an 'unused' volume to distributors back in China, which is where the company loses control over the final use of its memory chips. That said, calling YMTC directly responsible for where its memory devices end up may be a little bit of an over-reach. Meanwhile, YMTC owner Tsinghua Unigroup is a partially state-owned company, which is certainly a concern for the U.S. Ministry of Defence.

Recently, the Pentagon added YMTC to its list of concerns due to its possible links to the Chinese military. This list seeks to pinpoint companies that could jeopardize U.S. national security by aiding Beijing's military-industrial efforts.

While being on this list does not stop YMTC from operating in the U.S., it does block the company from getting contracts with the Defense Department. Moreover, this designation might prompt tougher actions such as being blacklisted by the U.S. Treasury Department, which would negatively impact their business activities in America. Yet again, YMTC hardly has any significant business activities in the U.S.  

Nonetheless, the inclusion into the list makes the life of the company harder, which is why it decided to make a statement claiming that it is not linked to the People's Republic's military activities.

The Pentagon recently designated over a dozen Chinese technology firms, including 3D NAND maker Yangtze Memory Technologies Corp. (YMTC), as 'military companies' that pose national security risks to the United States. The inclusion into the list does not cause any immediate effect, but it bans the U.S. military from buying devices featuring YTMC's chips, and it bars the companies from buying advanced chipmaking equipment based on U.S. intellectual property. Meanwhile, the Pentagon also excluded SMIC Hong Kong International Co Ltd from the list.

YMTC is among dozens of China-based companies included in the so-called 1260H list that highlights companies with alleged ties to the Chinese military and potentially supporting Beijing's military-industrial complex. YMTC is named to be a 'military company along with artificial intelligence (AI) firms Yitu Technology and Beijing Megvii, drone manufacturers Chengdu JOUAV and DJI Technology, lidar producer Hesai Technology, and tech company NetPosa. All of them have operations in the U.S.

Pentagon's 1260H list, established under the National Defense Authorization Act for 2021 and updated annually, serves as a mechanism for the Defense Department to flag companies that could threaten U.S. national security by allegedly supporting Beijing's military activities. The inclusion on the list does not entail a ban on operations within the U.S. but renders participants ineligible for Defense Department contracts. Moreover, being named can lead to further actions like blacklisting by the U.S. Treasury Department, which would significantly curtail their business prospects in the United States. 

An interesting thing about the current edition of the list is that the Pentagon excluded SMIC Hong Kong International Co. Ltd. from the list, citing that this unit does not belong to Semiconductor Manufacturing International Corp., China's largest foundry. Now, SMIC itself calls its Hong Kong unit a 'representative office.' That office naturally does not produce chips or is involved in any activities in mainland China. However, it is unclear whether the exclusion from the 1260H list opens SMIC doors to win contracts with the Pentagon.

SMIC and YMTC are on the U.S. Department of Commerce Bureau of Industry and Security's Entity List, and they cannot access advanced technologies developed in the USA, including wafer fab equipment. This is obviously a bigger concern for both companies than their ability to bid for U.S. government contracts.

Guangdong province in China has set up a new ¥11 billion ($1.5 billion) fund to boost its semiconductor industry, continuing China's regional chip fund trend. Shanghai started its chip fund back in 2016, and now has ¥28.5 billion ($4.02 billion) in it. These local efforts across China show how serious the country is about building a self-reliant chip industry — and they are all in line with the goals set by the federal government, according to a report by the South Central Morning Post. 

The new fund is called the Guangdong Semiconductor and Integrated Circuit Industry Equity Investment Fund Phase II. It follows the first phase, which launched in December 2020 and put ¥10 billion ($1.41 billion) into various chip projects in Guangdong. Both funds are designed to fund local semiconductor initiatives, although authorities of the region do not list projects that the funds have financed.

Money for the Guangdong fund comes mainly from the Guangdong government's Guangdong Yuecai Holdings, which owns over 90% of the fund. There's also cash coming in from Dongguan and Zhongshan, two cities that each have a 4.5% share. 

Guangdong's chip fund is a local version of something bigger: the China National Integrated Circuit Industry Investment Fund (the "Big Fund"). The Big Fund has been around since 2014 and has pumped loads of money into China's chip industry — despite some issues with corruption. It started with more than ¥138 billion ($19.462 billion), and grew even bigger in 2019 with over ¥200 billion ($28.2 billion) in its second phase. 

One of the peculiarities of Chinese funding of semiconductor projects is that the cash comes from multiple sources, including the Big Fund, local governments, and banks — some of which are government controlled. SMIC, China's foundry champion, for example, does not own many of the fabs that it operates. 

On one hand, such an approach greatly simplifies establishing new semiconductor fabrication facilities and ensures that companies such as SMIC are producing chips needed by local companies. On the other hand, this does alter SMIC's business model, which differs substantially from other foundries as the company does not have to follow typical CapEx depreciation schedules.

China's Yangtze Memory Technology Corp is readying its next-generation 3D NAND memory architecture — Xtacking 4.0 — despite severe sanctions against the company, according to a document seen by Tom's Hardware. The company has no plans to increase the number of layers with the two Xtacking 4.0 devices currently in development, but the family may get broader over time.

YMTC's Xtacking 4.0 lineup that the company has disclosed to some industry peers so far includes 128-layer X4-9060 3D TLC and 232-layer X4-9070 3D TLC NAND devices, which could eventually be used to build some of the best SSDs. The company plans to use string stacking for both. So technically, it will produce 3D NAND arrays with 64 and 116 active layers, which enables wafer fab equipment makers to keep supplying it with necessary tools without directly breaking U.S. export rules (so long as they obtain an export license from the U.S. Department of Commerce).

It is unclear precisely what advantages YMTC's Xtacking 4.0 memory will bring, but typically Yangtze Memory has increased data transfer rates and storage density with every new node. Given that the company faces severe restrictions and cannot obtain all the tools it needs, Xtacking 4.0 could somehow address its limitations. Plausible advances would include increasing the number of planes to boost parallelism and bitline/wordline optimizations to improve latency. As with any semiconductor, a tighter-tuned variant could also offer improved yields. 

YMTC has been producing its 232-layer 3D TLC NAND memory featuring its Xtacking 3.0 architecture in high volumes for over a year now and more recently expanded its Xtacking 3.0 family with 128-layer 3D TLC and 232-layer 3D QLC offerings that are cheaper, easier to make and which comply with restrictions set by the U.S. government for China. The company's Xtacking 3.0 family of products uses string stacking coupled with hybrid bonding that employs an older process node for the CMOS portion of the chip (similar to CuA, but with a separate die). 

Under the rules imposed by the U.S., Dutch, and Japanese governments, makers of wafer fab equipment should get an export license to sell equipment that can be used to make 3D NAND with 128 or more layers. Meanwhile, it looks like there is no formal limitation on stacking two or more wafers with less than 128 layers on top of each other. 

Technically, string stacking is not a wafer bonding technique per se, but from a manufacturing point of view, building a 3D NAND array on top of another 3D NAND array on one wafer is similar to producing two 3D NAND arrays on two separate wafers. Therefore, makers of equipment could be using this as yet another loophole to supply wafer fab tools to Chinese entities. 

Yangtze Memory Technologies (YMTC), China's largest 3D NAND memory maker, filed a lawsuit against Micron Technology for patent infringement on Thursday, November 9, in a California court. The legal battle involves eight U.S. patents related to 3D NAND technology that cover multiple basic aspects of how modern flash memory works and is manufactured. The lawsuit could be a part of YMTC's efforts to get removed from the U.S. government's blacklist, or it could be a part of the larger China-U.S. tensions. 

The lawsuit alleges that Micron used the patented tech, the cornerstone of YMTC's Xtacking technology, in its 128- and 176-layer 3D NAND products. Tthe lack of publicly available information for YMTC somewhat lowers the public view of the legal action (we discovered the lawsuit at TechNews.TW).

Here is the list of YMTC's patents that it alleges Micron has infringed:

• US10,950,623 covers methods for building 3D NAND memory chips, essential for the layered architecture that these chips utilize. 
• US11,501,822 focuses on the management of non-volatile storage devices, crucial for maintaining data without power. 
• US10,658,378 and US10,937,806 both deal with through-silicon via (TSV) technology, a fundamental tech for 3D NAND memory in general. 
• US10,861,872 describes a 3D NAND manufacturing technology.
• US11,468,957 covers architectural and operational aspects of NAND memory. 
• TUS11,600,342 is focused on methods for reading data from 3D NAND flash memory, another critical aspect of flash memory in general. 
• US10,868,031 is about techniques for stacking memory cells in multi-layered 3D NAND chips, another key aspect of 3D NAND in general.

The extensive nature of the lawsuit filed in California may be a sign of YMTC's backlash for inclusion it into the U.S. Entity List (which cost it an increase in CapEx around $7 billion), which severely cut the company's access to advanced wafer fab equipment. This legal action follows China's ban on Micron products in May over alleged cybersecurity concerns, though China has recently signaled a thaw in relations with Micron.

As with all patent disputes, the legal battle between YMTC and Micron could take years to work its way through the courts. YMTC's patents cover a wide range of basic NAND technology, but multiple companies, like Western Digital, SK hynix, Samsung, and Kioxia, also use similar architectures. This could make the patents difficult to defend in court. 

Although Yangtze Memory Technologies Co. (YMTC) has been mass-producing its 232-layer 3D TLC NAND for well over a year now, we rarely get to see this type of memory in action since it looks like it is mainly used in products that are sold in China. But perhaps we will finally experience it in a device that will be available globally. That device is Xiaomi's recently announced 14 Pro smartphone, reports BusinessKorea, and it may feature the world's highest-performing smartphone storage subsystem.

The Xiaomi 14 Pro is one of the most capable flagship smartphones released this year. The handset is based on Qualcomm's Snapdragon 8 Gen 3 application processor (one Cortex-X4 core at 3.30 GHz, five high-performance Cortex-A720 cores at 3.20 GHz, two Cortex-A520 at 2.30 GHz, and Adreno 750 graphics implemented using TSMC's N4P node) that is mated with 12 GB or 16 GB of LPDDR5X memory, and an up to 1 TB UFS 4.0 storage subsystem allegedly relying on YMTC's 232-layer 3D NAND memory with a 2400 MT/s interface. 

YMTC offers at least two products based on 232-layer 3D NAND production technology: its 1Tb X3-9070 3D TLC NAND device featuring six independent planes and its 1Tb 3D QLC NAND IC featuring four independent planes. Both feature the company's proprietary Xtacking 3.0 architecture and a 2400 MT/s interface, which makes them suitable for the best SSDs featuring a PCIe 4.0 or PCIe 5.0 interface. Meanwhile, the six-plane 3D TLC device is obviously faster than the quad-plane 3D QLC device. 

We do not know which YMTC's 232-layer 3D NAND devices are used by Xiaomi for its flagship 14 Pro smartphone (and we do not know if they are used because the information comes from an unofficial source), but a 16-bit NAND device with a 2400 MT/s interface speed can theoretically transfer up to 600 MB of data per second.

Now, since we are talking about 256 GB, 512 GB, and 1 TB UFS 4.0 storage subsystems (employing two, four, and eight 1Tb ICs), we may reasonably expect the first one to boast an up to 1.2 GB/s sequential read/write speed, the second one to offer an up to 2.40 GB/s sequential read/write speed, and the third one to provide up to 4.80 GB/s sequential read/write speed. The latter is the performance level of a good desktop-grade SSD, though we can only wonder in which use cases this level of performance can bring real-world benefits.

If the information about the usage of YMTC's 232-layer 3D NAND devices in the Xiaomi 14 Pro smartphone is correct, then the handset may well have the highest-performing storage subsystem in a smartphone.

Last week we reported that Changxin Xinqiao got ¥14.56 billion (approximately $1.99 billion) from the China Integrated Circuit Industry Investment Fund — the Big Fund. Bloomberg today reported that the company actually got considerably more money, which is an indicator that the Chinese government continues to invest in production of commodity memory, despite U.S. sanctions against the country's semiconductor sector. 

As it turns out, Changxin Xinqiao Memory Technologies raised ¥39 billion ($5.4 billion) from the state-backed China Integrated Circuit Industry Investment Fund Phase II and two additional investors affiliated with the local government from the eastern Chinese city of Hefei. The money will be spent on building a fab that will produce dynamic random access memory (DRAM).

All that said, neither $1.99 billion nor $5.4 billion is enough to build a leading-edge DRAM fab, since modern memory production facilities use expensive tools that are also used to make logic chips. It is enough to start building a fab, however, and having invested $5.4 billion, investors of the project will be less likely give up, so Changxin Xinqiao Memory Technologies will likely get its fab at some point.

It should be noted that Changxin Memory is managed by Zhao Lun, who happens to be chief executive of ChangXin Memory Technologies (CXMT), another DRAM maker funded by the Chinese government. Meanwhile, it is unclear whether the companies are affiliated or whether their investors expect the two companies to collaborate. For example, the two companies could share R&D expenses. 

The capital influx from the Big Fund is part of China's continued aim toward semiconductor self-sufficiency. Being the world's largest electronics manufacturing base, China consumes much more memory than it produces, so it makes a lot of sense to produce DRAM and 3D NAND memory in the country. But it remains to be seen whether these Chinese memory makers can produce competitive products, and at sufficient volume, given the fact that they cannot procure leading-edge wafer fab equipment from American, Dutch, and Japanese companies.

Yangtze Memory Technologies Corp (YMTC), China's largest maker of 3D NAND chips that are used to make flash memory for some of the best SSDs, recently secured a much-needed multi-billion dollar investment. This money comes after the company had to spend $7 billion in a year to overcome U.S. sanctions against the Chinese semiconductor sector and YMTC's inclusion on the Department of Commerce's Entity List.

Last year the company received a ¥50 billion ($7 billion) capital boost from its shareholders, which included contributions from the government-backed China Integrated Circuit Industry Investment Fund (the so-called Big Fund). But after the U.S. imposed restrictions on selling specific chipmaking tools to all Chinese companies in October and added YMTC to its blacklist, the company began bleeding financially — which is why it needed additional money from its investors. This new round of funding exceeded the company's expectations, according to a report from the Financial Times.

YMTC's cash reserves were substantially depleted as it grappled with the costs associated with repairing its equipment and procuring tools from third parties (where possible). YMTC has been actively seeking to reduce its reliance on American chipmaking by turning to alternative suppliers from Europe, Japan, and South Korea. Just like other Chinese companies, YMTC is attempting to increase its reliance on Chinese suppliers — namely Advanced Micro-Fabrication Equipment (AMEC) and Naura.

"If Chinese companies have equipment that can be used, [YMTC] will use it," an investor told FT. "If not, it will see if countries other than the U.S. can sell to it. If that does not work, YMTC will develop it together with the supplier."

AMEC and Naura can produce quite competitive etching equipment, but it is unclear whether they can produce it in large volumes. To keep developing production nodes that are more advanced than the company's leading-edge 232-layer 3D NAND, YMTC will need wafer fab equipment that is more advanced than the tools currently made by AMEC and Naura. According to the FT report, these tools will have to be co-developed with YMTC. 

It is unclear where YMTC can get advanced lithography tools from, given the existing sanctions against Chinese chip sector by Japan and the Netherlands. The best lithography tools that China's industry is about to deliver later this year can make chips on a 28nm-class node, but that is not enough for advanced logic or 3D NAND.

YMTC chairman Chen Nanxiang, who was recently appointed as the head of the China Semiconductor Industry Association, has called for solidarity and cooperation within the industry. This call to unity is a response to what Chen describes as an 'unprecedented upheaval' in the global supply chain, a sentiment that reflects the Chinese industrial perspective on both the current geopolitical tech tensions in general, as well as the specific sanctions against the Chinese semiconductor industry.

China-based Yangtze Memory Technologies (YMTC) has quietly started shipments of 3D QLC NAND memory with 232 active layers, TechInsights has discovered. The new 3D NAND memory devices boast the industry's highest recording density of as well as extreme performance due to Xtacking 3.0 architecture.

YMTC's 1Tb 3D QLC NAND device has a recording density of 19.8 Gbit/mm2, which is the world's highest density for a commercial IC. In fact, even a 232-layer 3D TLC NAND chip from YMTC boasts a recording density of 15.47 Gbit/mm2, which is higher than competing offerings that are in mass production.

There is a substantial difference between the 3D QLC and 3D TLC NAND devices from YMTC: The former features a quad-plane design to optimize die size, whereas the latter features a six-plane design to maximize performance. Meanwhile, both ICs use Xtacking 3.0 architecture as well as a 2400 MT/s data transfer rate, so both could be used for the best SSDs featuring a PCIe 4.0 or PCIe 5.0 interface.

One interesting thing to note is that analysts from TechInsights discovered YMTC's 1Tb 3D QLC NAND devices in a ZhiTai Ti600 1TB solid state drive, which was quietly launched in July, 2023, and has been on the market since then. YMTC has been producing this memory for over a quarter now. Yet, it is unclear whether it can produce such memory in substantial quantities.

Producing such memory devices is a major achievement for YMTC, which is blacklisted by the U.S. government and cannot obtain leading-edge tools from American companies. Meanwhile, the company has managed to achieve the highest recording density with its 3D TLC NAND with 232 active layers, surpassing its rivalsprimarily due to its Xtacking 3.0 architecture that relies on hybrid bonding.

Yangtze Memory Technologies (YMTC) has teamed up with China-based SSD brands and developers of flash controllers to market solid-state drives based on its 232-layer 3D NAND memory, reports DigiTimes. While the company does not formally have access to wafer fab equipment required to make such flash memory and cannot sell such devices to multinational drive suppliers, it can still sell in China. 

YMTC has traditionally worked with Maxio Technology, a China-based developer of SSD controllers, to design solid-state drive platforms. So it is not particularly surprising that the two companies have a platform based on 232-layer 3D NAND. DigiTimes claims that YMTC has teamed up with other China-based SSD controller firms to build drives featuring its latest memory and their chips, though it does not disclose the names of these companies.

One SSD supplier that markets 232-layer 3D TLC NAND-based drives is Tengyn. The manufacturer started to sell SSDs featuring YMTC's X3-9070 memory a couple of months ago, and since then, their prices have dropped sharply. Like in other parts of the world, there is an oversupply of 3D NAND in China, so low SSD prices are expected. Meanwhile, since YMTC has trouble selling its advanced 3D NAND memory elsewhere, it is probably willing to sell it at a discount in the People's Republic to keep its fabs busy.

YMTC's 232-layer 3D TLC NAND memory featuring its Xtacking 3.0 architecture was developed to address applications like the best client and server SSDs needing high performance and storage density. The memory devices feature a 15.47Gb/mm2 areal density, according to Yole Group, and a 2400 MT/s interface. 3D NAND devices featuring such a speedy I/O can be used to build PCIe Gen5 x4 drives that offer throughput of over 12 GB/s, and such drives are still not common.

Yet, because the U.S. government blacklists YMTC, international SSD makers are reluctant to use its advanced 3D NAND memory for their products due to possible risks.

The Chinese company Yangtze Memory Technologies (YMTC) seems to be adapting to new US sanctions realities with a rather mysterious 3D NAND product that combines a relatively low storage density with an extremely speedy interface. In addition, YMTC has once again raised prices for its 128-layer 3D NAND by about 5%, according to DigiTimes. This move comes despite the ongoing slump in the PC and consumer electronics markets in general and the 3D NAND memory market in particular and follows a previous increase in May.

Mysterious New Product

Although stringent U.S. export rules have hindered YMTC's ability to produce 3D NAND flash chips with more than 128 layers, the company is swiftly adopting alternative strategies. In particular, YMTC has initiated large-scale production of its 128-layer NAND flash chips, dubbed 'Wudangshan,' that reportedly relies on the Xtacking 3.0 architecture, according to the report.

YMTC has productized a version of 128-layer flash before with its Xtacking 2.0 architecture, but the step up to the 3.0 revision marks a significant overhaul of an existing product, which normally wouldn't take place. If true, the US sanctions, which hinder YMTC's ability to make flash with higher layer counts, have spurred the company to re-vamp its older flash with performance-boosting tech. If such 3D NAND devices exist, they may combine relatively low storage density with high performance, which could be the company's stop-gap solution before it comes up with a product that offers both high capacity and high performance. 

The key point of Xtacking is to produce a 3D NAND array and various peripheral logic on separate wafers using the most efficient memory and logic process technologies in YMTC's possession. Xtacking 3.0 introduces back-side source connect (BSSC) for the memory cell wafer, which simplifies production flow and reduces costs, according to TechInsights. YMTC developed Xtacking 3.0 for its 232-layer 3D NAND to bring together ultimate storage density (15.47Gb/mm2, according to Yole Group) and a 2400 MT/s interface, therefore offering perhaps the world's best 3D NAND.

 It is theoretically possible to add a CMOS wafer with high-speed peripheral logic to a 128-layer 3D NAND array wafer, but the question is whether it brings benefits for real-world applications because there is only so much one can do with a 128-layer process technology and 3D NAND device capacity from a costs perspective.

In fact, based on Yole Group's findings, YMTC's 232-layer 3D NAND uses string stacking to achieve 232 layers, with 128 layers on the 'first deck' and 125 layers on the 'second deck'. From a manufacturing point of view, this means that the company may simply not add the 'second deck' to get to a 128-layer 3D NAND featuring the Xtacking 3.0 architecture. But, of course, this severely hurts the storage density of the final device and its competitive position.

Yet, it could make sense for the company to produce such memory if it wants to address specific applications and/or as a stopgap before it finds a way to add the 'second deck' without using equipment that it cannot procure or service.

YMTC has not formally confirmed the very existence of 128-layer 3D NAND with Xtacking 3.0 codenamed Wudangshan, so take the information with a grain of salt though.

5% Price Hike

YMTC recently raised its unit price for a 512Gb NAND device from $1.5 to $1.6, following an earlier hike from $1.35 - $1.4 in May. This pricing strategy contrasts with international NAND suppliers who are grappling with attempts to raise prices amidst tepid customer demand. But Yangtze Memory seems to be taking advantage of its position now that Chinese government agencies cannot buy memory from Micron.

DigiTimes claims that thanks to substantial financial backing from the China IC Industry Investment Fund (the Big Fund), YMTC is seeking solutions for upstream equipment and materials. While the report does not divulge whether YMTC is seeking to use domestically produced wafer fab equipment, it is certainly a possibility.

Apparently, the company's stockpile of consumables, accumulated before the sanctions, is anticipated to sustain the mass production of 128-layer 3D NAND products for at least three more years, the report says. Meanwhile, it remains to be seen whether 128-layer 3D NAND will still be competitive three years down the road. Of course, the Chinese government may force certain organizations to only use YMTC's 3D NAND for their systems, which will guarantee stable demand, but it remains to be seen whether that demand will be significant enough to support the company economically.

Chen Nanxiang, the head of Chinese 3D NAND champion Yangtze Memory Technologies (YMTC) that got into the U.S. trade blacklist late last year, asked producers of wafer fab equipment to buy back tools the company cannot use. In his keynote speech at Semicon China conference he also warned that the global semiconductor industry is entering an era of turmoil and disarray due to increasing geopolitical tensions.  

Buy It Back

According to Bloomberg, Chen Nanxiang, chairman and CEO of YMTC, gave a speech in Shanghai, China, that directly addressed machine suppliers, and how they should buy back the legally purchased equipment and components if they cannot be used.

While many companies are slowing down spending on wafer fab equipment these days, makers of chip production equipment have huge backlogs, so selling off fab tools may not be a problem for YMTC, if it really wants to get rid of them.

YMTC got hit by the U.S. government twice last year. In October, YMTC was blocked from acquiring wafer fab equipment (WFE) with American technology necessary for producing 3D NAND with 128 or more layers as part of the broad sanctions imposed by the U.S. government on the Chinese semiconductor industry. The restrictions also oblige U.S. citizens to obtain a license from the U.S. Department of Commerce to support development or production of said memory devices in China. As a result, four prominent chipmaking tool companies — ASML, Applied Materials, KLA, and Lam Research — ceased their work with YMTC due to the requirement of obtaining relevant export licenses from the DoC.

In December, YMTC got into the U.S. DoC Entity List and can no longer procure equipment, software, and other technologies from U.S.-based companies (or other companies selling technologies developed in America) — unless the latter secure a special export license from the DoC. Such export licenses are reviewed with a presumption of denial.

As a result of the two hits, YMTC cannot obtain, install, and use the latest tools it acquired from WFE producers that it needed to make 232-layer 3D NAND, a type of memory that could power some of the best SSDs with a PCIe 5.0 interface. Furthermore, the company has troubles servicing its existing tools. Some analysts believe that Yangtze Memory might have to abandon production of 3D NAND and turn itself into maker of specialty 2D NAND or even a contract chipmaker.

Turmoil and Disorder

Chen stated that globalization is effectively dead, and the previously established rules, balance, and harmony have been disrupted. This shift is predicted to greatly impact the supply chain, industry practices, and business models. The head of YTMC believes that key aspects of globalization such as competition, innovation, and the free movement of talent and resources have been affected. 

According to reports via Nikkei, Chen believes that globalization is dead and that government interventions and that existing rules and past balances are now broken. Chen continues, stating that the semiconductor industry could see a period of turmoil. A period which could see a tremendous and profound impact on the supply chain and industry practices.

The chief executive and chairman of the troubled 3D NAND maker highlighted the importance of global cooperation in the semiconductor industry. He emphasized that the complex chip supply chain involves 25 countries and has indirect ties with another 23. 

Chen also indicated that semiconductor markets in China and the U.S. — which benefited from contributions from foreign investors and inventors — do not solely belong to their respective countries, but are global assets. Similarly, he brought example of NAND flash memory, which was invented by Toshiba in Japan back in the 1980s, and which is perfected by South Korea-based Samsung and SK Hynix.

The goals of China's 'Made in China 2025' program cannot be achieved, according to a report by DigiTimes that cites Wei Shaojun, a professor at Tsinghua University and Deputy Chairman of the China Semiconductor Industry Association. Not only will Chinese chipmakers not be able to satisfy 70% of the local IC demand in 2025, but their profitability, capitalization, and ability to invest in development are greatly behind those of their rivals.

In a recent address at the China Nansha International IC Industry Forum, Wei Shaojun pointed out that despite a rise in the value share of domestically produced chips from 13% in 2013 to 41.4% in 2022, the ambitious goal of reaching a 70% self-reliance rate for semiconductors, a key objective of the "Made in China 2025" initiative, seems nearly unattainable. That's due to the current global shifts that made both U.S. and Europe fund their semiconductor sectors from the governments' pocket.

Wei highlighted an imbalance between supply and demand in China's IC design sector, pointing out that while the monthly demand of the country is about 1.5 million 300mm wafer starts per month (WSPM), domestically owned Chinese semiconductor producers can only provide a monthly output of 440,000 WSPMs.

The professor from Tsinghua University, an organization that has been particularly instrumental in the development of the local IC sector, underscored a critical misapprehension regarding the advancement of China's chip fabrication sector over the last ten years.

The accelerated growth of China's semiconductor production industry was largely due to foreign companies operating within the country. From 2016 onward, Chinese investor-owned semiconductor companies have experienced an average compounded annual growth rate (CAGR) of 14.7%. Yet, non-Chinese wafer fabrication enterprises from Taiwan, South Korea, and elsewhere, have seen a higher CAGR of 30%. This not only doubles the expansion pace of Chinese-owned businesses but also underlines the continued reliance on external assistance in China's semiconductor manufacturing industry.

The highly ranked executive of the China Semiconductor Industry Association also noted that the Chinese publicly traded semiconductor companies are currently underperforming, marked by the IC design sector's low gross profit margins.

Amid the AI boom that skyrocketed Nvidia's market value to over $1 trillion, Wei Shaojun pointed out that 135 Chinese semiconductor firms on China's STAR and ChiNext markets have a combined value of less than half of Nvidia's. Moreover, these firms' average gross profit margin of 39.1% in 2022, particularly 34.2% for the STAR-listed IC design companies, is significantly lower than the over 60% enjoyed by U.S. chip developers indicating that Chinese companies are still operating at lower profitability levels.

Judging by what we've seen from China's domestic CPU and GPU efforts, there's a lot of work remaining before it will close the gap.

The Made in China 2025 national strategic plan to develop the domestic semiconductor industry inspired the establishment of tens of thousands of fabless chip designers in the People's Republic in just a few years. However, it seems many of them were unfit to survive intense competition between each other and global players, which is why about 10,000 such chip companies went out of business during 2021–2022, reports DigiTimes.

Market observers now wonder whether the massive shuttering of Chinese IC design companies is a result of tighter U.S. export control imposed in 2020–2022, or the ongoing global semiconductor industry downturn. While both factors contributed to the closures, there were a number of issues specific to China that drove around 10,000 homegrown chip companies out of business.

The Made in China 2025 program implemented several policies to meet its goals, including lower taxes for high-tech companies, encouraging acquisition of foreign tech company, supporting R&D funding by large manufacturers, and direct state R&D funding, among other things. The results were inspiring to say the least. The number of Chinese chip developers increased from 736 in 2015 to 1,780 in 2017, according to China Renaissance Securities. Then in 2020–2021 as many as 70,000 chip companies were registered, according to DigiTimes.

The report admits that the escalation of the U.S.-China trade war in 2018 further inspired the Chinese government to fund high-tech companies, with many of the IC design houses established due to subsidies from federal or local governments. In addition to subsidies, a rush of speculative capital drove the establishment of similar IC design firms developing commodity chips that were poised to crash for multiple reasons.

The semiconductor industry is well known for being capital intensive, but the sector is heavily dependent on talent, skillful management, and knowledge. Even if a company has ample resources, it still needs to attract talent, invest in research and development, and ensure sufficient production capacity from foundries for sustainable long-term development. Without engineering talent and proper management, the chances of success are not high. While there are loads of engineers in China these days, there are not enough managers to run these IC designers successfully.

Meanwhile, an investment bubble emerged. Ding Xing Quantum, a private equity firm based in China, has been investing in domestic IC designers since 2017. The company observed that at the onset, the worth of such a company ranged between RMB 200–300 million ($28–$43 million). However, by 2019, the valuation of startup companies in this sector ballooned to over RMB 1–2 billion ($145–$190 million), which clearly points to an investment bubble, and such bubbles tend to crash.

There was another factor that contributed to the shortfall of Chinese IC design companies. Underperformance of the Chinese consumer market transitioned into a structural imbalance of supply and demand starting Q3 2021, and then the global semiconductor industry faced inventory correction in the second half of 2022 and entered its own downturn. As a result, the demand for chips dropped in general, and China-based developers that produced commodity ICs went bankrupt since they could not offer anything special.

The sanctions against the Chinese semiconductor industry clearly affected development of the sector, as it is evident from the problems faced by companies like Alibaba, Biren, HiSilicon, and YMTC. Meanwhile, the global semiconductor downturn and the inability of many Chinese chip designers to compete played a far bigger role in the dissolution of 10,000 semiconductor entities in the People's Republic, the story by DigiTimes concludes.

Yangtze Memory Technologies Corp (YMTC) won't be able to produce competitive 3D NAND memory in a couple of years after being placed on the U.S. Commerce Department's Entity List, says TrendForce. Following the sanctions, it will become difficult for the company to procure wafer fab equipment (WFE) and other goods from U.S.-based companies, hampering its bit growth. On the other hand, its partners outside of China may cease working with YMTC. 

Due to the blacklist, the U.S. DoC will review all individual transactions of wafer fab equipment, software, technologies, support services, and other goods of U.S. origin to the Chinese 3D NAND maker. Such export license applications are reviewed with the presumption of denial, so every procurement process is becoming tough and prolonged for YMTC.  

This will have a particularly drastic effect on YMTC's bit output increase as it involves procuring loads of advanced tools from various companies, including those from the U.S., Japan, and the Netherlands. Since getting new equipment from the U.S. is now arduous, YMTC will be severely constrained with its 3D NAND output increase. Furthermore, suppose Japan and the Netherlands follow the U.S. with restrictions against the Chinese semiconductor industry. In that case, it will get even more challenging for YMTC and other companies in the People's Republic to procure WFE from foreign companies. 

TrendForce initially believed that YMTC's 232-layer 3D NAND memory ramp featuring the Xtacking 3.0 architecture and other measures would uplift the company's bit output by 60% in 2023 compared to 2022. However, after the U.S. government imposed sweeping sanctions against Chinese chipmakers in October, it revised its forecast to an 18% increase. Now that YMTC is blacklisted, TrendForce believes its bit output will decrease by 7% year-over-year in 2023. 

Since it will be close to impossible for YMTC to procure leading-edge fab tools and other technologies from American companies, YMTC will lose its competitiveness due to technological stagnation, TrendForce asserts. By contrast, other 3D NAND makers will have transitioned to 200-layer 3D NAND memory by 2024 and might even start making leading-edge 300-layer 3D NAND products for the best SSDs by that time.  

If YMTC does not keep up with its peers with a 232-layer 3D NAND ramp, it will lose market share as flash memory from companies like Micron, SK Hynix, and Samsung will be more cost-effective.  

Another reason why YMTC is set to lose market share is that companies from outside of China may prefer not to work with a company in the U.S. DoC's Entity List. TrendForce claims that P.C. OEMs planning to qualify YMTC's client SSDs have 'temporarily halted the customer sampling and adoption process.' As a result, analysts believe that YMTC will be forced to focus solely on its domestic market. 

TrendForce thinks that to survive, YMTC might need to transition to making specialty 2D NAND flash or even transform into a contract maker of chips focusing on mature fabrication processes. 

The U.S. government is set to add China-based 3D NAND maker YMTC to the Department of Commerce's "Entity List" this week, along with dozens of other high-tech companies from the People's Republic, according to a report by DigiTimes. Companies on the Entity List will not be able to procure equipment, software, and other technologies from American companies— unless the latter obtain a special export license from the DoC. 

As a maker of 3D NAND, YMTC already faced problems created by the U.S. government's sweeping sanctions against Chinese semiconductor sector. YMTC is already unable to procure American wafer fab equipment (WFE) to make 3D NAND with 128 or more layers. As a result, four leading U.S. makers of chipmaking tools have already stopped working with YMTC, as they must get appropriate export licenses from the Department of Commerce.  

Being placed on the Entity List creates further difficulties for YMTC, as it will now also lose access to all kinds of wafer fab equipment that is made in the U.S. or that contains American IP (including tools for inspection), software, and other technologies.  

The Bureau of Industry and Security (BIS) of the U.S. Department of Commerce added Yangtze Memory Technologies Company along with 30 other entities to the Unverified List (UVL) in early October as it could not identify bona fides (end users) of their products and whether said bona fides were involved in boosting China's military capabilities. 

Once a company is placed on the UVL, it has 60 days to prove its products do not break any export regulation rules. This usually means the company must allow the U.S. DoC to conduct inspections and verifications. If the checks are unable to be completed to the U.S. DoC's satisfaction, the company is then placed on the Entity List.

Although the Chinese government appeared to relent last week, allowing U.S. export control to inspect several entities (including YMTC), these checks take some time. It looks like they haven't been completed yet, and so YMTC and the other companies will be placed on the Entity List for the time being. 

The Entity List is essentially a blacklist, and getting blacklisted is a big deal, because YMTC will lose access to all American technologies subject to regulations. When Huawei and its subsidiaries were placed on the Entity List, they lost access to software and hardware that used any technology designed in the U.S., including (but not limited to) electronic design automation (EDA) software used for chip design, as well as chips produced by TSMC. This significantly constrained Huawei's ability to develop its system-on-chips, and almost eliminated its ability to make them in volume. 

Now, YMTC and 30 other Chinese companies will face the same issues as Huawei — and only time will tell whether all of them will be able to survive without access to American technology. 

YMTC has begun shipments of its Zhitai TiPlus7100 SSDs based on its latest 232-layer 3D NAND Xtacking 3.0 memory with a 2400 MT/s interface, ITHome reports. This confirms that YMTC's latest 3D NAND memory — which is required to produce SSDs with a PCIe 5.0 x4 interface that fully saturate this interface and hit a 12.4 GB/s sequential read speed — is now in mass production.  

YMTC's Zhitai TiPlus7100 are M.2-2280 PCIe 4.0 x4 drives designed to combine affordability with strong performance. The SSDs are to be available in 512GB, 1TB, and 2TB versions with the speediest rated for a sequential read speed of up to 7000 MB/s as well as a sequential write speed of up to 6000 MB/s. 

When it comes to random performance, then 1TB and 2TB models are set to offer up to 900K random read IOPS as well as up to 700K random write IOPS. The TiPlus7100 drives do not carry any SDRAM buffer and uses a host memory buffer, which is an indicator that we are dealing with a reasonably priced product. Meanwhile, these drives can easily challenge the best SSDs available today. 

YMTC does not disclose what controller it uses for its Zhitai TiPlus7100 SSDs, but the main point about these drives for us is that they use the company's 1Tb X3-9070 chips — 232-layer six-plane 3D TLC NAND memory devices with a 2400 MT/s interface and the company's proprietary Xtacking 3.0 architecture.  

The 1Tb X3-9070 device not only boasts a bit density of 15.03 Gb/mm^2 (as revealed by TechInsights), which by far exceeds the bit density of 1Tb 3D TLC NAND memory ICs with less than 200 layers, but it also features an ultra-fast 2400 MT/s interface.

Earlier this week, Micron introduced its Micron 2550 drives based on its 232-layer six-plane 3D TLC NAND devices that is said to have a 14.6 Gb/mm^2 bit density, which outstrips YMTC's 232-layer 3D TLC ICs in terms of bit density. Meanwhile, Micron's ICs currently shipped have a 1600 MT/s interface, which is good enough for mainstream drives, but not good enough for ultra-high-performance SSDs with a PCIe 5.0 x4 interface.  

Therefore, while YMTC is not the only company to mass produce 3D NAND with over 200 layers, it is the first company that is mass-producing memory with a 2400 MT/s I/O. This is not going to last long though as Micron plans in initiate production of 232-layer 3D NAND with a 2400 MT/s interface in early 2023.

The U.S. administration says that the Chinese government allowed U.S. export control inspections of People's Republic's high-tech companies that were placed on its Unverified List in October, according to a report from the Financial Times. This move seems designed to ensure that Chinese 3D NAND company YMTC does not get placed on the Department of Commerce's "Entity List," which would severely damage the company's ability to procure equipment from American companies.  

YMTC and a number of Chinese high-tech companies were placed on the Unverified List in early October because the U.S. government could not verify whether their products (or products made using their products) ended up in the hands of China's military. Once a company is placed on the Unverified List, it has 60 days to prove its products do not break any rules. If the company cannot prove this within 60 days, it is placed on a trade blacklist called the "Entity List" and, in the case of YMTC, is denied the use of any American technology.

Normally, the Chinese government refuses to allow U.S. export controls inspectors access to domestic companies. However, the government made an exception in the case of YMTC and some other firms from Tianxia, most likely because YMTC, and the semiconductor industry in general, is so essential for the country. 

"We are seeing better behavior," Alan Estevez, the US commerce under-secretary for industry and security, said to the Financial Times. "Mofcom has been more forthcoming. […] We are seeing a change in attitude. It’s not the first time we’ve seen a such a change in attitude, so it depends on how long that is sustained."

It's too early to guess what the U.S. export controls inspectors will find, but at least YMTC won't be blacklisted — for now. That said, four leading U.S. makers of wafer fab equipment have already stopped working with YMTC due to the latest export rules imposed by the U.S. government in October, as they needed to get appropriate export licenses from the Department of Commerce. 

If YMTC doesn't get blacklisted and can prove it doesn't supply products for military use, perhaps it will be easier for companies like Applied Materials, KLA, and Lam Research to obtain approval from the U.S. government to continue working with the 3D NAND maker. 

While in general Chinese chipmakers struggle to successfully rival established players, there are a few exceptions. YMTC (Yangtze Memory Technology Co.) is perhaps the most widely known one, a 3D NAND memory manufacturer that successfully competes against offerings from well-known producers. But to become competitive, The New York Times reports that YMTC received support from an unexpected ally, Apple.

Apple assisted YMTC in hiring engineers from established Western companies in order to improve its yields and productivity, according to the NYT report that cites three people familiar with the matter. So far, neither Apple nor YMTC have confirmed or denied the information, though the California-based consumer electronics giant is known for helping its manufacturing partners to build their operations.

YMTC's 3D NAND devices use the company's unique Xtacking architecture. That consists of two separate wafers bonded together to build ultra-dense and ultra-fast flash memory devices that can power the best SSDs.

With Xtacking, one wafer is used to produce NAND memory arrays leveraging the most efficient 3D NAND process technology that YMTC can design. Then a second wafer is used for various peripheral logic made on the company's most efficient logic process technology. Once the memory arrays and logic are connected using metal TSVs (through silicon vias), the resulting NAND can offer the best of both worlds: ultra-dense memory arrays and a very fast interface.

YMTC's rivals produce memory arrays and peripheral logic on the same wafer, which makes it harder for them to push their interface speeds to the limits. That in turn constrains performance of client-grade PCIe 5.0 x4 SSDs.

Making 3D NAND the way YMTC does is somewhat harder than making memory using traditional methods employed by Kioxia and Western Digital, Samsung, SK Hynix, and Micron. YMTC's 3D NAND suffered from low yields (according to media reports) and a slow ramp up, so apparently Apple helped the Chinese companies to hire engineers from reputable manufacturers to fix its issues.

For Apple, which uses 3D NAND across the vast majority of its products, it's important to diversify its supply as it helps to get better prices from each manufacturer. Helping YMTC to ramp up production means there will be more 3D NAND devices to choose from, at lower prices. Meanwhile, now that the U.S. administration has imposed severe sanctions against Chinse semiconductor and supercomputer industries, it looks like Apple no longer plans to use YMTC's 3D NAND even for its products bound to be sold in China.

YMTC became widely known by the general public after it was included in the U.S. Department of Commerce's Unverified List in early October. The U.S. DoC's UVL includes entities whose bona fides (end users) could not be identified "satisfactorily for reasons outside the U.S. Government's control." If YMTC cannot prove to the DoC that its memory is not used by the Chinese military or security forces within 60 days after inclusion into the list, the DoC can include the company in its entity list. That would require American companies to obtain a special export license to sell their products to YMTC.

Meanwhile, four leading makers of wafer fab equipment have already ceased working with YMTC due to the latest export rules imposed by the U.S. government in October. We've reached out to Apple and YMTC for comment and will update if we receive any further details.

Yangtze Memory Technologies Co. (YMTC) on Friday denied that it met with Chinese government officials to discuss the impact of U.S. sanctions against supercomputer and semiconductor industries. The company called the information false and rumors but neither confirmed nor denied the meetings took place.  

YMTC claims that it never participated in such closed-door meetings and the information about such participation damages its reputation. 

"The above remarks about our company published by relevant media are not true," YMTC said in a statement. "In the face of rumors, YMTC has always insisted that those who clean up will themselves clean up. However, in the current situation of major changes in the market environment and industrial environment, the above-mentioned sinister and false statements will not only damage the corporate image of YMTC, but also have a serious negative impact on the semiconductor industry environment at home and abroad." 

Bloomberg reported on Thursday that China's Ministry of Industry and Information Technology (MIIT) had summoned executives from the country's leading semiconductor and supercomputing companies, including Sugon and YMTC, to discuss the effects of sweeping U.S. sanctions against Chinese high-tech industry on their businesses. The agency said that the government promised to support affected companies financially but did not present any plans for how the effects of the sanctions that essentially bar China from the global semiconductor supply chain can be relieved. 

While we cannot verify whether the closed-door emergency meetings took place and whether YMTC executives participated in them, we can make some guesses about the situation based on the publicly available details. 

YMTC is a subsidiary of government-controlled Tsinghua Unigroup, founded in collaboration between Tsinghua Unigroup, China's national 'Big Fund' aimed to develop the semiconductor industry in the country, and Hubei provincial government. Given YMTC's close ties with government-controlled organizations, the MIIT could hold a meeting with executives of Tsinghua Unigroup, who could talk on behalf of its subsidiary to the government. However, in this scenario, YMTC executives did not need to participate in any meetings. 

While this assumption makes sense, YMTC would prefer not to be associated with the Chinese government (or rather with the Chinese Communist Party) in any way at the moment. Although Yangtze Memory must secure wafer fab equipment from American companies to keep operating, it will also have to procure more advanced fab tools to ramp up production of its next-generation 3D NAND memory featuring the Xtacking 3.0 architecture with ~200 layers and a 2400 MT/s interface speed. 

Meanwhile, to sell equipment to YMTC, U.S.-based companies like Applied Materials, KLA, and Lam Research will need to get an appropriate export license from the U.S. Department of Commerce, which will review applications with the presumption of denial. To that end, any publicity that associates YMTC with the People's Republic's government (which means CCP) is damaging for the company.

"YMTC a commercial company that adheres to the concept of globalization, marketization and compliance operation," the statement from YMTC reads. "Since its establishment, it has always adhered to the principle of legal and compliant operation worldwide. In the future, we will also adhere to the original aspiration and adhere to the principle, and actively seek the path of development under the current changes in the external environment and situation." 

Besides refuting participation in closed-door meetings with government officials and stating that it is a commercial company that adheres to principles of globalization, YMTC also threatened to take legal action against publications, journalists, organizations, and individuals who reported, re-reported, or forwarded the original Bloomberg story.

Four leading makers of chip production equipment have ceased sales of new tools and provide support services of installed tools to China-based Yangtze Memory Technologies Co. (YMTC) due to new export rules that came into effect on Wednesday. To sell new equipment to YMTC and support existing machinery, the toolmakers must get a technology export license from the Department of Commerce’s Bureau of Industry and Security (BIS). 

U.S.-based Applied Materials, KLA, and Lam Research have ceased supplying equipment to produce 3D NAND memory with 128-layers or more to YMTC under the latest export rules that impose new license requirements for semiconductor production equipment destined for China starting October 12. However, they will continue to supply new equipment and tools to multinational corporations that produce chips in China — Samsung, SK Hynix, and TSMC — for a year without a license. Yet, all export license applications to supply tools to Tsinghua Unigroup-controlled YMTC require review with a presumption of denial. Tsinghua Unigroup is a government-controlled organization.

For the same reason, ASML, which is in the Netherlands, told its U.S. employees to stop ‘servicing, shipping or providing support to any customers in China until further notice,’ according to a Bloomberg report. 

While Applied, KLA, and Lam Research ship metrology, etching, deposition, inspection, and die sorting tools to Yangtze Memory, ASML sells the Chinese 3D NAND maker crucially important lithographic scanners. While formally, lithography tools used by YMTC are not as advanced as tools used by logic or dynamic random access memory (DRAM) producers and theoretically should not fall under the new licensing requirements, ASML prefers to assess the new export rules and ensure that it complies with the new regulations. 

YMTC attempted to procure all equipment and spare parts it could from its American partners in China in the recent weeks and days to keep producing 3D NAND without disruptions to meet the demands of its clients, which reportedly include Apple. While it probably acquired all manufacturing equipment and spare parts it could (although nobody has confirmed or denied this), its main challenge now is to install and deploy new tools and parts. The new export restrictions imposed by the U.S. government also include services. As a result, ASML, Applied Materials, KLA, and Lam Research employees cannot assist YMTC, as well as other Chinese chipmakers, which means that they cannot install new tools or replace faulty components. 

Whether YMTC’s engineers at its fab are qualified to service equipment produced by companies from the U.S. and the Netherlands is unknown. However, for now, the 3D NAND bit supply from YMTC is secure, according to DigiTimes. 

Meanwhile, if YMTC’s employees cannot deploy new tools and fix production tools, sooner or later, YMTC will have to scale down or even cease production of 3D NAND memory. The latter is unlikely to happen as the company will struggle to survive. Still, without support from toolmakers, it will get much harder to maintain YMTC’s production facilities. 

It is noteworthy that YMTC was getting ready to ramp up its next-generation family of six-plane 3D NAND chips featuring the company’s Xtacking 3.0 architecture with ~200 layers and a 2400 MT/s interface speed. These chips could eventually enable some of the best SSDs, but we are not sure whether YMTC will be able to get the right tools to produce such memory devices.

Taiwanese semiconductor manufacturing prowess might be one reason for China to invade the island and seize fabs belonging to Taiwan Semiconductor Manufacturing Co., United Microelectronics Corp., and Micron. One of the potential responses to such a plan could be evacuating personnel and destroying the fabs, suggested Parameters, a top U.S. army publication recently. But this might be unnecessary, according to Taiwan’s National Security Bureau.

To build chips using leading-edge process technologies, TSMC needs leading-edge chip production equipment from companies like ASML, Applied Materials, and KLA. Even if China invades the island and seizes TSMC’s fabs without access to advanced equipment and ultra-pure raw materials, it will be impossible for China not only to keep developing leading-edge manufacturing nodes but to keep production on current technologies uninterrupted.

“TSMC needs to integrate global elements before producing high-end chips,” Chen Ming-tong, director-general of Taiwan’s National Security Bureau, told Taiwanese lawmakers this week, reports Bloomberg. “Without components or equipment like ASML’s lithography equipment, without any key components, there is no way TSMC can continue its production. […] Even if China got a hold of the golden hen, it won’t be able to lay golden eggs.”

China’s slowing economy, tensions with the U.S., and internal political battles in recent years increased the probability of China’s invasion of Taiwan to seize the island, get a hold of multiple world-class technologies, and improve the approval rating of Xi Jinping. But probability does not mean certainty. China must maintain relationships with the United States and the European Union, its two key trade partners. Furthermore, without access to manufacturing tools and technologies designed in America and Europe and without money from trade partners, China’s occupation of Taiwan might turn into a Pyrrhic victory.

New Export Rules

The U.S. sanctions against China’s supercomputer and semiconductor sectors prove relatively efficient. Late last week, the Department of Commerce’s Bureau of Industry and Security (BIS) published new export rules that impose new license requirements for semiconductor production equipment destined for China starting October 12.

Under the new rules, U.S. companies must obtain a license from the U.S. DoC for tools that can make logic chips using 14nm/16nm nodes or thinner, DRAM ICs on nodes of 18nm and below, and 3D NAND chips with 128 layers or more. Licenses for fabs owned by Chinese entities will face a ‘presumption of denial,’ and licenses held by multinational corporations will be decided case by case.

U.S.-based Applied Materials, KLA, and Lam Research this week ceased to supply appropriate tools to their clients in China, including Semiconductor Manufacturing International Corp. (SMIC) and Yangtze Memory (YMTC). Also, the said companies started to withdraw their employees from YMTC’s fabs. The decision has already reduced the value of the global semiconductor sector by hundreds of billions of dollars, and it remains to be seen how significantly it will affect the businesses of Applied, KLA, and Lam Research.

Yet, the move shows how the U.S. can crack down on China’s semiconductor industry in a few days.

The U.S. administration is trying to restrict China's military potential by denying access to government-controlled entities leading processor designs, chip development, and production capabilities, according to Bloomberg. As a result, it wipes billions of dollars from the semiconductor industry's global market value. Some U.S. companies are already restricting access to certain China-based companies to their products and services, whereas others are waiting before the U.S. government makes its restrictions official.

Being the world's second-largest economy and the global factory, China unsurprisingly consumes loads of semiconductors and produces loads of chips in the country. China must adopt the latest technologies designed in the country and elsewhere to support its economic and humanitarian development. Still, those technologies could also help advance Chinese military potential, something that nobody in the region and outside of it likes. Advanced weapons will boost China's influence on the economy and geopolitics, which poses risks not only to U.S. allies like Japan, South Korea, and Taiwan but also to the U.S.

To crack down on advancements in China's military development or at least slow it down tangibly, the U.S. has been gradually imposing restrictions on the Chinese chip industry for several years. Still, the Biden administration is ready to impose even more severe sanctions against China's chip sector in the coming weeks or months. In particular, it looks like the current government wants Chinese contract chipmaker Semiconductor Manufacturing International Co. (SMIC) to stop making chips using its 14nm-class fabrication technology (and thinner) nodes as well as Yangtze Memory Technologies Co. (YMTC) to cease advancing its 3D NAND memory.

The U.S. authorities plan to impose stricter rules on exports of advanced U.S.-originated semiconductor technologies to China. But because the chip industry is global, decisions of the U.S. administration will impact the whole semiconductor supply chain. Capitalizations of large chip companies have been decreasing for a while now as investors have been expecting slowing demand for chips (and it is here). With more restrictions from the U.S., more semiconductor companies will suffer. As a result, the Philadelphia Stock Exchange Semiconductor Index — a modified market capitalization-weighted index composed of semiconductor companies — hit its low this week.

Semiconductor Companies Set to Suffer

Several types of semiconductor industry companies will suffer from the U.S. attempting to crack down Chinese high-tech industry to prevent the development of the country's military potential.

Chip development begins with intellectual property and electronic design automation (EDA) tools. Companies like Ansys, Cadence, Keysight, and Synopsys sell plenty of software and IP blocks to Chinese chip designers. Approximately 13% of Cadence's and Synopsys' 2021 revenue came from China, and that does not count the various chip designs developed elsewhere with China in mind. There are other providers of EDA and similar tools from other countries, and while Cadency and Synopsys are market leaders, usage of software from Siemens EDA (Germany), Zuken (Japan), or Primarius Technologies (China) for chip designs is possible.

We are not sure whether the American government has the power to restrict licensing of Arm chip designs to customers in China. Still, potentially it could confine licensing of certain portions of designs developed in the U.S. Also, the U.S. administration can restrict the usage of American chip development tools to create chips for certain Chinese entities by contract chip designers like Alchip.

Foundries. Hundreds of chip designers in China develop chips requiring sub-14nm/16nm process technologies. At least a dozen companies could use leading-edge production nodes like TSMC's N4/N5 or Samsung's 3GAE/5LPP. If they are prohibited from using advanced manufacturing processes (including relatively mainstream 14nm/16nm-class nodes), contract makers of chips will lose revenue and profits.

Makers of semiconductor production equipment. China is not exactly a leading nation in chip production. Still, with thousands of chip developers in the country (most working on ICs for emerging applications like Internet-of-Things, smart homes, smart cities, etc.), it needs loads of semiconductor production capacity in the country. As a result, China is one of the largest markets for companies like ASML, Applied Materials, KLA, and Lam Research.

Even without the U.S. administration announcing stricter restraints against China's memory and logic production companies, KLA on Tuesday stopped sales of specific tools to Chinese companies that can produce 3D NAND with 128 layers or more, DRAM chips at an 18nm node or thinner, and advanced logic chips, according to Reuters. Despite expectations, KLA stopped shipments to companies like SMIC and YMTC and companies like SK Hynix that produce memory in China.

Chip designers like AMD, Nvidia, and Intel also sell boatloads of advanced chips to various Chinese companies. However, recently the U.S. government barred sales of AMD's, and Nvidia's high-performance compute GPUs to any customers in China, including cloud companies like Alibaba and Baidu, which could have lowered Nvidia's quarterly revenue by up to $400 million if the government did not allow the company to fulfill existing orders and continue to work with Chinese manufacturing partners for a while.

In particular, the U.S. government does not want American companies to sell Chinese entities supercomputing-grade hardware that could allow building machines with the performance of over 100 FP64 PetaFLOPS or over 200 FP32 PetaFLOPS within 41,600 cubic feet (1178 cubic meters), according to China Renaissance.

Some say that the decision by the U.S. administration will speed up the development of high-end compute GPUs by Chinese companies like Biren Technologies; others note that Nvidia is years ahead of any China-based firm regarding software stack for artificial intelligence and high-performance computing. But in any case, China is a large market for high-performance CPUs and GPUs, so the crackdown will inevitably hurt companies like AMD, Nvidia, and Intel.

Will it send Chinese cloud giants back or contain the development of new military capabilities is up for debate. On the one hand, it will hurt companies like Alibaba, Baidu, and Tencent. But on the other hand, they will still be able to buy less capable hardware from AMD and Nvidia, so they will not lose access to American technologies overnight. Still, they will have to spend more on electricity, data center space, and maintenance.

Will It Work?

The U.S. sanctions against China's high-tech sector aim to control what the country gets and restrict some of the technologies to the People's Republic, not wholly destroying its high-technology and economic potential. Therefore, the appropriate U.S. organizations approved 88% of tech exports to China in 2021.

But if the percentage of approvals gets dramatically lower, Chinese semiconductor companies will suffer (and the whole industry with it). China will take years and hundreds of billions of dollars to replace all technologies with U.S. origins.

On the chip production side of matters, the good news for China is that Japan-based chip tools companies like Nikon, Canon, Tokyo Electron, and several others will suffer from the U.S. crackdown and will likely attempt to develop equipment that does not use technology designed in the USA. However, assuming that they succeed and China-based makers will make progress in tools, Chinese SMIC and YMTC will be able to continue advancing their chip technologies (assuming that the U.S. government does not try to curb the production of chips using trailing nodes).

On the chip design side of matters, things are considerably more complex. Local chip developers barely have experience with the design of ultra-large supercomputer-grade chips. While some items may be mitigated by exploiting a modern chiplet-based design approach, there are not enough engineers or senior managers in China to address all large-scale semiconductor projects that the country might need to replace things like Nvidia's A100/H100 compute GPUs, according to China Renaissance. Meanwhile, Taiwan and the U.S. now restrict the hiring of chip specialists by Chinese companies, so getting the right talent will be harder for Chinese companies.

While severe restrictions for the Chinese semiconductor industry might have a drastic effect on local chipmakers, the U.S. government has not entirely restricted access of Chinese companies to American technologies, which has enabled companies like Biren to develop their world-class chips. However, if this happens, the effect will be far-reaching. It will take years for China to replace technologies that originated in the U.S., which will dramatically affect the country's economic development.

But so far, investors' worries about a significant chip demand slump and fears about the future of the global semiconductor industry if the U.S. decides to impose stricter sanctions against the Chinese chip sector have erased more capitalization than any concrete actions by the U.S. administration.

The U.S. Department of Commerce on Friday added 31 companies from China to its Unverified List (UVL), a step that precedes the addition to the Entity List. Among the entities added to the UVL is Yangtze Memory Technologies Co. and various companies involved in Chinese semiconductor and adjacent high-tech industries. Inclusion to the UVL adds roadblocks for companies to acquire technologies from the U.S. 

The U.S. DoC's UVL includes entities whose bona fides (end users) could not be identified "satisfactorily for reasons outside the U.S. Government's control." Inclusion to the UVL means that U.S. suppliers with ties to such companies will have to conduct additional due diligence to verify end users of final products before shipping their tools to entities from the list, reports Reuters. This might potentially mean they will have to apply for additional licenses. Meanwhile, unlike inclusion to the DoC's Entity List, inclusion into the UVL does not mean that U.S. companies will have to get special export licenses for all the goods they ship to the listed companies. 

The inclusion of YMTC into the UVL means that U.S.-based producers of chipmaking tools and software must perform additional checks and notify the U.S. government before shipping their equipment to the 3D NAND memory maker. This does not completely cut off YMTC from new fab tools but makes it harder for the company to procure equipment from companies like Applied Materials, KLA, or Lam Research. Furthermore, some licenses will likely not be granted. 

"The use of license exceptions for exports, reexports, and transfers (in-country) involving a party or parties to the transaction who are listed on the UVL is suspended," a statement of the DoC reads. "Additionally, […] there is a requirement for exporters, re-exporters, and transferors to obtain (and maintain a record of) a UVL statement from a party or parties to the transaction who are listed on the UVL before proceeding with exports, reexports, and transfers (in-country) to such persons, when the exports, reexports and transfers (in-country) are not subject to a license requirement. Finally, […] Electronic Export Information (EEI) must be filed in the Automated Export System (AES) for all exports of tangible items subject to the EAR when a party or parties to the transaction is/are listed on the UVL."

U.S. officials have been talking about restricting YMTC's access to technologies with American origins for weeks now, but the government has not yet blacklisted the company.  

The Biden administration is also considering toughening up access of China-based semiconductor companies to leading-edge equipment used to make chips on 14nm/16nm-class production technologies. This will not disrupt supplies of advanced tools to Samsung and SK Hynix that make 3D NAND and DRAM chips in China, according to a Reuters report, but will substantially affect the competitive positions of companies like Semiconductor Manufacturing International Co. or Yangtze Memory. 

The Semiconductor Industry Association (SIA) has yet to assess how the new export control rules affect U.S.-based chipmaking tools companies. 

"We are assessing the impact of the new export controls on the U.S. semiconductor industry and working with our member companies and the U.S. government to ensure compliance," a statement by the SIA reads. "We understand the goal of ensuring national security and urge the U.S. government to implement the rules in a targeted way — and in collaboration with international partners — to help level the playing field and mitigate unintended harm to U.S. innovation."

Adding another 3D NAND maker to its supply chain is certainly something that is good for Apple, but U.S. legislators are not particularly happy that one of the world's largest consumer electronics companies will use memory from YMTC and thereby essentially help China's semiconductor industry to develop. Apple argues that it plans to use chips from Yangtze Memory only for iPhones to be sold in China. 

Apple has been evaluating the use of 3D NAND from YMTC for months now, and last week a report emerged that the company finally decided to use flash memory from the Chinese manufacturer for its iPhones. The company confirmed to Financial Times that it has been considering using YMTC's products for some of its smartphones sold in China, yet it did not confirm that these chips were used as of last week. There were no plans to use YMTC's devices in smartphones sold outside of China, Apple is reported to have said. But the U.S. lawmakers are still concerned about Apple working with Yangtze Memory. 

YMTC is largely controlled by Tsinghua Unigroup, a company directly to the Chinese government via various banks and investment funds. It also has ties with Tsinghua University, funded by the Ministry of Education. As a result, YMTC is to a large degree  a state-controlled company, which is a concern of U.S. legislators. In fact, they accuse the Chinese government of illegally subsidizing YMTC, giving it a competitive edge over American makers of memory, such as Micron. This of course ignores the $76 billion Chips bill.

Another reason why U.S. lawmakers are worried about YMTC is that it allegedly supplied 3D NAND memory and/or products to Huawei without obtaining an export license from the U.S. Department of Commerce. As of 2020, everyone making products for Huawei using technologies developed in the U.S. (hardware tools used in semiconductor fabs, electronic design automation tools used to develop chips, etc.) must obtain a license from the U.S. Department of Commerce.

"YMTC has extensive ties to the Chinese Communist party and military," said Michael McCaul, a congressman for Texas who sits in the House foreign affairs committee, in an interview with Financial Times. "There is credible evidence that YMTC is breaking export control laws by selling goods to Huawei. Apple will effectively be transferring knowledge and know-how to YMTC that will supercharge its capabilities and help the CCP achieve its national goals." 

Some lawmakers want commerce secretary Gina Raimondo to put YMTC on the entity list for its ties with CCP, and therefore ban it from getting U.S. tools and software, according to the Financial Times's sources.

While YMTC is a very competitive company that indeed has close ties with the Chinese government, it is only one of Apple's suppliers from China. There are tens of companies from China that supply parts for Apple's iPhones and other products. Actually, even some OLED screens for Apple's smartphones are supplied by Beijing-based BOE Technology Group. That said, even without YMTC, Apple's products are full of technology developed and made in China. That might change eventually as Apple shifts some of its production to India and other countries, but that will of course be a slow process. 

Apple has reportedly approved 3D NAND flash produced by Yangtze Memory Technologies Corp. and will now use it for its upcoming iPhone 14 smartphones. The move represents a strong win for YMTC and will ensure steady supply of flash memory for Apple's next-generation products. 

For now, Yangtze Memory will supply Apple 3D NAND for its upcoming iPhone 14 handsets, reports BusinessKorea. Smartphones are Apple's most popular products, so Apple needs boatload of DRAM and NAND for its smartphones. But eventually Apple could adopt YMTC's memory for other products as well. For example, YMTC has extremely competitive products incoming, such as its latest family of six-plane 3D NAND chips featuring the company's Xtacking 3.0 architecture and a 2400 MT/s interface speed. These chips could eventually enable some of the best SSDs. 

It is going to take a while for YMTC's latest products to mature and get into Apple's other products. But considering specifications offered by the company's latest 3D NAND devices as well as Apple's expertise in 3D NAND and controllers, YMTC's have all chances to land into iPads or Macs at some point. 

Being one of the world's largest consumers of 3D NAND flash, Apple tends to procure memory from different vendors, including Kioxia, Samsung, and SK Hynix. Adding YMTC to the supply chain means that Apple will now have more choice from characteristics and performance points of view as well as better positions to negotiate prices as many products supplied by the four manufacturers are more or less identical. 

From YMTC standpoint, winning a supply contract with Apple is a big deal as so far the company has primarily shipped its 3D NAND as well as solid-state drives to Chinese vendors. 

Just like other makers of commodity 3D NAND, Yangtze Memory has been evolving from a manufacturer of flash memory into a vertically integrated producer of solid-state storage solutions. The company started with simplistic SATA SSDs roughly two years ago, then evolved to more advanced PCIe drives, and now it is time for YMTC to enter the most lucrative 3D NAND market — enterprise storage.

YMTC's first enterprise-grade product is the PE310 SSD that comes in a 2.5-inch U.2 form factor featuring a PCIe 4.0 x4 interface, using 3D TLC NAND memory, and offering storage capacities of up to 6.4TB for drives aimed at mixed workloads and designed for up to three drive writes per day (DWPD) as well as up to 7.68TB for drives targeted at read-intensive workloads and designed for up to 1 DWPD.

Regarding performance, the drives offer up to 6.2 GBps sequential read speed, up to 4.5 GBps sequential write speed, up to 1 million read IOPS, and up to 380K/190K (mixed workloads/read-intensive workloads) write IOPS. Since Phison's PE310 are enterprise-grade products, stated performance numbers represent sustained, not peak, performance. While the drives are very fast, since they are enterprise-oriented, we will not include them in our list of the best SSDs available today.

Yangtze Memory's 3D TLC NAND memory used for PE310 drives features Xtacking 2.0 architecture. It, therefore, supports a maximum interface data transfer speed of up to 1,600 MT/s, which is in line with what the latest 3D NAND ICs from makers like Samsung and SK Hynix offer.

YMTC does not disclose which controller it uses for its PE310-series products, just like it did not a few months ago. The only thing Yangtze Memory says is that controllers of its PE310-series drive feature LDPC ECC, are compliant with an NVMe protocol, and support AES-256 encryption. In addition, the drives support such functions as power loss protection, advanced eight-level power management, end-to-end data protection (E2EDP), thermal throttle, and so on.

Considering that YMTC intentionally does not disclose almost anything about PE310's controller, including the supported NVMe version or branding of LDPC-based signal processing and ECC technology, the company either ultimately does not want to reveal the off-controller it uses (in 2020, the company teamed up with Phison for client SSDs, but it does not automatically matter that it uses chips from the said vendor on the PE310), uses a highly-customized version of an off-the-shelf controller that has little to do with the initial chip, or uses an in-house developed controller.

Developing an enterprise-grade solid-state drive is complicated as YMTC has to guarantee ultimate reliability, very high endurance, consistent performance, and predictable power consumption under high loads. However, as it turns out, the company has managed to do this. As a result, it will now compete against solutions from other vertically integrated SSD solutions providers, such as Kioxia, Micron, Samsung, and Western Digital.

While the U.S. semiconductor developers and production tools producers welcomed the U.S. CHIPS Act and Science law that will provide subsidies to American chip developers and manufacturers, two Chinese trade associations believe that the legislation hinders innovation, discriminates against foreign companies, and intensifies geopolitical tensions.

"The act will intensify the global geopolitical competition in the semiconductor sector and hinder global economic recovery and future technology innovations," a joint statement by the China Council for the Promotion of International Trade (CCPIT) and China Chamber of International Commerce (CCOIC) published by the state-backed Global Times reads.

The recently signed CHIPS Act and Science law authorizes the U.S. government to award local chipmakers about $52 billion in grants and other incentives as well as provide a 25% investment tax credit for new fabs worth up to $24 billion. Meanwhile, companies who receive funds under the CHIPS Act and Science legislation will be prohibited from building new or expanding existing fabs in China and other countries that present a threat to the U.S. national security over the following 10 years. In addition, the new law will grant tens of billions of dollars to various high-tech research and development projects in the U.S.

"On the one hand, it is a typical industry-specific subsidy, which does not conform to the non-discrimination principle of the World Trade Organization," the statement partly published by ChinaDaily reads. "On the other hand, the bill identifies particular countries as key targets, which leads companies to be forced to adjust their global development strategies and layouts."

It is noteworthy that China's local authorities co-own fabs operated by Semiconductor Manufacturing International Corp. (SMIC), the country's largest contract maker of chips. Meanwhile, state government backs Tsinghua Unigroup, which happens to own YMTC, the only national 3D NAND producer, as well as controls tens of chip designers.

"The US is using government power to forcibly change the international division of labor in the semiconductor field and harm the interests of companies from all over the world, including Chinese and American companies," a quote published by Anadolu Agency reads.

Nowadays the U.S. only produces 12% of the global chip output and is significantly behind Taiwan, South Korea, and China. Subsidies for chipmakers in the U.S. will make it more appealing for them to build fabs in America rather than in Asian countries. But another important part of the legislation promotes research and development in the U.S. About half of the world's chips are designed in the U.S., with additional subsidies from the federal government that share could grow, something that Chinese trade groups seem to also oppose.

In addition to subsidizing semiconductor research, development, and production in the U.S., the U.S. government has been restricting sales of American chip production equipment to Chinese manufacturers (e.g., SMIC) to prevent the country from using American technologies and equipment in military purposes. Meanwhile, CCPIT and CCOIC are confident that U.S. restrictions will not slow down Chinese semiconductor industry development.

In recent years the U.S. government has taken steps to crackdown on Chinese production of logic chips using advanced fabrication processes. In a bid to prevent the country from advancing its military capabilities. In addition to logic chips, U.S. officials are now mulling a crackdown on China's 3D NAND production industry. The biggest issue is that in addition to domestic makers, such restrictions will affect Samsung and SK Hynix. 

The U.S. government is considering banning shipments of American 3D NAND production equipment to manufacturers of flash memory with more than 128 layers in China, which will affect Yangtze Memory Technologies Co. (YMTC), China's only domestic 3D NAND maker, reports Reuters citing sources familiar with the matter. But in addition, such a move will also hurt Samsung and SK Hynix, which produce 3D NAND memory in China too. 

At present China's YMTC commands about 5% of global 3D NAND production, whereas Samsung, SK Hynix, and YMTC produce about 23% of the global 3D NAND output in China, according to Yole Intelligence cited by Reuters. 

Previously the U.S. government restricted shipments of advanced chip production tools to China citing national security concerns. By contrast, 3D NAND hardly poses any threat to national security. Meanwhile, 3D NAND made in China competes against 3D NAND produced by American companies — Micron Technology and Western Digital. Protecting them from cheap flash memory produced in China will improve their market positions, but will naturally upset South Korea-based Samsung and SK Hynix. 

If the action is approved, this will not only disrupt production of 3D NAND in China and leave the market without crucially important commodity memory, but will also hurt U.S.-based companies like Applied Materials, KLA, and Lam Research, which sell large amounts of equipment to China.  

While American companies produce crucially important fab tools, there are companies from Japan, South Korea, and even Taiwan that may substitute some of the equipment made in the U.S. As a result, it is unclear whether the U.S. ban will have a desired effect on Chinese 3D NAND production industry in the long term. 

Earlier this week the U.S. government imposed new export rules under which American companies will have to apply for an export license if they sell semiconductor production tools capable of producing chips using 14nm-class and thinner nodes.

CORRECTION (5/17/2022): YMTC says that the PE310 and PE320 are not controllers but the model numbers of SSDs it produces. The company is not disclosing what controllers these drives use and who makes them so we have no way of knowing whether the company is in-fact developing its controllers or not. The original article text is below.

--------------

All leading makers of 3D NAND memory have their in-house developed controllers for solid-state drives that power their own-brand SSDs. Being a considerably smaller 3D NAND manufacturer, Yangtze Memory Technologies Co. is following in the steps of its larger rivals by designing its own SSD controllers. Recently two of these controllers passed PCI Express compliance validation at a PCI-SIG workshop.

YMTC's family of PCI-SIG-validated SSD controllers currently include models PE310 and PE320, though there may be other controllers that are not listed in the PCI-SIG Integrators List. Both controllers feature a PCIe 4.0 x4 host interface and are designed for higher-end drives. Since both chips have the same host interface, we can speculate that they support a differing number of NAND flash channels. 

Speaking of NAND channels, it is necessary to note that one advantage that YMTC's 3D NAND chips have over rivals is their ultra-fast interface. This benefit is enabled by the company's Xtacking technology, which entails manufacturing a 3D NAND memory array and a NAND logic (address decoding, page buffers, etc.) on two different wafers and then bonding them together. YMTC once said that its Xtacking could enable I/O speeds of 3 GT/s (up from 1.6 ~ 2.0 GT/s data transfer rate on today's 3D NAND chips from companies like Samsung or SK Hynix), so building a controller that supports this transfer rate could incredibly fast SSDs (that may even end up in our list of the best SSDs available today).

Historically, among makers of 2D and then 3D NAND, only Samsung and Intel designed SSD controllers in-house, whereas other leading manufacturers of flash memory left the development of SSD controllers to third parties. In many cases, lousy combinations of NAND, controller, and firmware led to inferior drives.  

But as the SSD market grew from tens of millions of units in the late 2000s to hundreds of millions of units in 2015 and 3D NAND and SSD architectures got more complex, all makers of NAND memory (except Toshiba, which owned a stake in Phison) initiated the development of SSD controllers in-house. These are two key reasons why makers of memory started to develop their own SSD controllers:

• New generations of 3D NAND memory require very advanced error correction and signal processing algorithms. For obvious reasons, developers of 3D NAND devices can bring together a particular set of algorithms to a particular 3D NAND IC better than a third party. Also, new generations of SSDs require more sophisticated bad block management, end-to-end data path protection, advanced security functions and many other technologies that have to be prudently implemented, keeping capabilities of particular 3D NAND ICs in mind.
• Meticulous integration of 3D NAND memory, controller, and firmware allows making fast and reliable drives, something that consumers and OEMs demand. Selling tens of millions of SSD is far more profitable than selling millions of 3D NAND wafers or tons 3D NAND chips.

YMTC clearly has similar motives as other suppliers of 3D NAND memory, so the development of own-brand controllers (or even ordering custom controllers from a third party and then enabling exclusive features) is an entirely logical and expected move from a Chinese 3D NAND producer. 

YMTC previously collaborated with Phison and Silicon Motion to enable turnkey SSD solutions and its own-brand drives. We fully expect the company to continue working with both controller developers in the future. In fact, Samsung is the only maker of 3D NAND that does not sell its SSD-graded memory on the open market. Other manufacturers supply their chips to various SSD makers, which is why they also have to work with SMI and Phison to ensure compatibility with controllers.

Apple is reportedly considering adding China-based Yangtze Memory Technologies Co. (YMTC) to its list of 3D NAND memory producers to diversify its supply chain and protect against sudden interruptions. YMTC would hardly become a major Apple supplier, but adoption of its memory by one of the world's largest makers of electronics will be a huge step forward for the Chinese semiconductor industry.

Apple Needs More 3D NAND

Apple has been negotiating with YMTC it for months, but the agreement has not been reached yet, according to a Bloomberg report that cites sources with knowledge of the matter. After Kioxia and Western Digital lost 6.5 exabytes of 3D NAND flash due to contamination issues at one of their fabs in Japan and increased risk of supply disruptions, Apple got more inclined to add another 3D NAND supplier to its list. Apple is now reportedly evaluating 3D NAND samples from YMTC.

Apple uses 3D NAND flash for all of its devices. Since the company designs controllers and firmware for solid-state storage in house, it can tailor them for memory from different manufacturers.

Yangtze Memory is the only China-based developer of 3D NAND memory that produces chips in volume. The company is considerably smaller than Apple's suppliers Kioxia, Micron, Samsung, and others, so it will hardly challenge these manufacturers. In fact, given how much 3D NAND flash Apple uses for its devices (and the fact that sales of the company's PCs, smartphones, and tablets are growing, just as content of NAND flash inside them), it remains to be seen whether and how quickly YMTC will be able to meet Apple's demands in terms of volume. 

More Suppliers from China

Virtually all of Apple's PCs, smartphones, tablets, and other products are made in China by contract manufacturers like Foxconn Electronics or Pegatron. Yet, the majority of components Apple's products use come from outside of China. But Tianxia is slowly gaining share inside devices designed by the California-based company. 

Apple already uses displays from BOE Technology Group (former Beijing Oriental Electronics), a major Chinese maker of display panels. Furthermore, the company most probably uses certain chips produced by Semiconductor Manufacturing International Corp. (SMIC) for its clients, who are Apple suppliers. Therefore, adoption of YMTC's 3D NAND will be another milestone in Apple's adding of Chinese manufacturers to its roster of suppliers.

Can YMTC Supply Enough 3D NAND for Apple?

Apple is reportedly a very demanding customer, placing large orders and reportedly expecting priority. YMTC's actual abilities to supply to Apple and its existing customers is something that remains to be seen.

YMTC is a crown jewel in a holding controlled by troubled Tsinghua Unigroup, which is backed by the state, which is undergoing restructuring. Back in 2019, YMTC had plans to expand its 3D NAND production capacity to around 150,000 3D NAND wafer starts per months (WSPMs), but financial difficulties of its parent most probably affected these intentions. As of early 2021,  the company's production capacity was about 100,000 64-layer 3D NAND WSPM, according to Nikkei.

By now, Yangtze Memory has started production of 128-layer 3D NAND memory, which uses more process steps, wafers spend more time in chemical vapor deposition (CVD) machines, so the number of WSPMs gets lower but the number of NAND memory bits gets higher. It should also be kept in mind that unlike other makers of 3D NAND, YMTC produces NAND memory array and a NAND logic (address decoding, page buffers, etc.) on two different wafers (and at two different fabs) and then connects the memory arrays to the logic by metal via its proprietary Xtacking technology. 

So while YMTC has two 300-mm semiconductor production facilities that can house plenty of tools, its actual production capacity is unknown. Furthermore, it is unknown whether the company can rapidly expand its manufacturing capabilities if it needs to.

A team of researchers with the Tsinghua University in Beijing, China, has produced the world's smallest transistor. The research paper, published in Nature, describes how the new transistor design leverages an atom-thin graphene sheet as the transistor's gate, enabling a record-setting gate length of 0.34 nm. The team expects its research - and the resulting sidewall transistors, as they've named the result - to provide a way out from the greatly exaggerated news surrounding the death of Moore's Law. 

Transistors are the fundamental units of semiconductor design, requiring three building blocks to be operable: a source (where the electrical current enters the transistor); a drain (the point where that same electrical current leaves the transistor); and a gate, which controls whether or not electrical current makes that particular voyage. However, a transistor by itself is so simple that it isn't actually capable of any useful work. To that end, transistors are crammed together in integrated circuits; a certain number and arrangement of transistors (in scales of billions) then results in a CPU core, while another results in a SRAM cache bank, or a GPU.

It follows that one way to increase performance and available workload complexity lies in cramming ever more transistors into the same space. Another would be to increase the functionality of each transistor. All semiconductor manufacturing improvements are ultimately reaching towards the same: a higher number of transistors, however complex they are, per squared inch. And this is where millions of dollars worth of research are funneled into each year: looking for ways to decrease the size of transistors, so that more of them can be packed in the same area. That in itself is pretty straightforward; however, there are additional benefits to smaller transistors - since electrical current has to endure a smaller journey from the transistor's source towards its drain, it remains in the transistor's features for smaller amounts of time - simultaneously improving power efficiency, temperatures and operational frequency while reducing leakages.

But how did the researchers achieve such a small gate length? By taking advantage of the fact that graphene sheets, made out of graphene atoms laid out in a 2D plane, are only as thick as the graphene atom. That's how the 0.34 nm gate length was achieved: it corresponds to the height of the graphene sheet.

The classic transistor design was also changed to accommodate this radically new gate (of which the researchers built working prototypes). Imagine two adjacent buildings with different heights. In the transistor design, these buildings are made out of layers of silicon and silicon dioxide (insulators, meaning that they don't carry electrical currents). The graphene layer, one atom thick, is applied on top of the tallest building. But since graphene is electrically conductive (which is why we're using it in the first place), it has to be insulated on both sides; hence, the researchers applied another layer of insulating material: aluminum oxide. 

This leaves the very edges of the graphene layer (our building) as the only non-insulated elements, which confers the gate's one-atom thickness. And because the silicon "buildings" are at different heights, we can now apply a layer of semiconductor molybdenum disulfide over both and alongside the sidewall (hence, the name) of the tallest building. It's through this layer that electrical current will flow through the transistor. But remember that the atom-thick layer of graphene is making contact with the molybdenum disulfide electrical highway, which allows it to act as a gate, permitting (or interrupting) the current flow.

Finally, a source is placed on the tallest (silicon) building for electrical current to enter the molybdenum disulfide highway; and a drain, through which electrical current abandons the transistor, is placed on top of the other. That is if the gate (the graphene layer) allows for it to reach its destination. We now have a working transistor. 

Crucially, the researchers say these new transistors are easy to make with currently implemented technology and that the change in structure still resulted in compact units. In fact, they should ultimately prove less complex to produce than at least some other currently explored transistor designs since the placement of the layers of graphene and molybdenum disulfide is a relatively low-precision task (compared to etching and other semiconductor manufacturing techniques).

We should always keep in mind that not all promising research turns into a working product, and those that do (including this new transistor design) will take years to reach us. Even so, it's important to have a clear path ahead - or at least, fraught with options - than no option at all. This is especially true given the rate at which the worldwide semiconductor market is developing. That, paired with the increase in computational requirements from general computing, AI, Big Data, Web3, blockchain, and all other technological products, will demand extreme efficiency gains. Having a transistor feature as small as an atom certainly sounds apt for the aim.

According to a report from Nikkei, debt-stricken Tsinghua Unigroup will scrap plans to build two new large 3D NAND and DRAM fabs and reshuffle Unisoc as part of its restructuring plan. Both moves represent major setbacks for China's 'Made in China 2025' semiconductor self-reliance plan.  

Tsinghua Unigroup will axe two major semiconductor fab projects in China as part of the ongoing restructuring procedures, a serious blow for the country's semiconductor self-sufficiency goal as local memory production is a key part of the plan. The move will impact the global supply of 3D NAND and DRAM and thus affect memory prices two or three years from now. In addition, Tsinghua will reorganize Unisoc, the country's largest developer of a mobile system-on-chips (SoCs), which might be good news for MediaTek and Qualcomm but another blow for the self-sufficiency plan.

New Fabs Scrapped

Tsinghua Unigroup already owns Yangtze Memory Technologies Co. (YMTC), a rapidly developing yet still fairly small 3D NAND producer. The company announced intentions to build a 3D NAND and a DRAM fab worth $30 billion in 2017. Still, the initiative has been reconsidered several times (the locations of the fabs were changed, the amount of money to be invested was projected to increased) and has never materialized. After Tsinghua Unigroup faced major financial difficulties last year, the destiny of both projects was completely uncertain, especially amid tensions between China and the U.S. 

Tsinghua Unigroup's second 3D NAND project was essentially meant to replicate YMTC's success. Yet, since YMTC is yet to become a rival for Samsung Semiconductor and SK Hynix — two South Korean companies that produce 3D NAND in China — it did not really make much sense to build yet another fab, which could cost as much as $24 billion, reports Nikkei. 

To build its DRAM operation, Tsinghua hired Yukio Sakamoto, a former chief executive of Elpida Memory who has experience competing against companies like Samsung and Micron. However, Mr. Sakamoto left Tsinghua in the second half of last year after the company faced bankruptcy, even before he could use his knowledge and connections. Since it would have taken years and cost billions of dollars to develop a competitive DRAM process technology amid uncertainties with the supply of fab tools in 2025 ~ 2026, Tsinghua reportedly axed its DRAM intentions, too.

Unisoc to Be Restructured

Unisoc, a smartphone and tablet SoC developer, is yet another Tsinghua Unigroup business unit to be restructured. The company, which consolidates assets of consolidated Spreadtrum and RDA Microelectronics, reportedly controls about 9% of application processors for mobile devices globally. Still, since it mostly sells inexpensive 4G devices, it barely brings in huge profits and cannot compete against SoC designers with 5G-supporting products.  

The current plan includes further consolidation and relocation of existing engineers from multiple sites across China. While this might increase efficiency due to better collaboration, it will also postpone the realization of ongoing projects, possibly including the development of 5G SoCs. 

New Strategic Investors

With many assets under its control and huge plans for expansion, Tsinghua Unigroup owed approximately $31 billion and had $8 billion in cash and cash equivalents as of mid-2020. Yet it had defaulted or cross-defaulted on seven bonds worth $3.6 billion by the end of 2020, according to Reuters. As a result, Hong Kong-listed state-owned Huishang Bank requested the First Intermediate People's Court of Beijing Municipality to start insolvency procedures for the conglomerate in 2021. 

Late last year, two new investors — Beijing Jianguang Asset Management Co. and Wise Road Capital — emerged as new backers of Tsinghua Unigroup. Nikkei believes that the two funds acquired a portion of the stake controlled by Beijing Jiankun Investment Group, a private investment group founded by Zhao Weiguo, the chairman of Tsinghua Unigroup. 

Both Beijing Jianguang Asset Management Co. (also known as JAC Capital) and Wise Road Capital are keen high-tech investors. Over the past several years, they acquired several semiconductor companies, including chipmaker Nexperia, once a subsidiary of NXP Semiconductor, and Chinese chip packaging facilities of ASE Technology Holding. Last year they attempted to buy Magnachip Semiconductor but were denied by the U.S. and South Korean governments. 

With new investors on board, Tsinghua Unigroup will be more financially stable than before, but scrapping two major semiconductor projects emphasizes that the goals of the 'Made in China 2025' project are not going to be achieved. 

Tsinghua Unigroup, a Chinese high-tech conglomerate, on Friday confirmed that it had received a notification from a Beijing court informing it that one of its creditors demanded insolvency protection of the indebted company. As of mid-2020, Tsinghua Unigroup owed approximately $31 billion and had $8 billion in cash and cash equivalents, according to Reuters.

Tsinghua Unigroup is the owner of numerous Chinese semiconductor companies, including YMTC, the country's only 3D NAND maker, and Unisoc, a mobile SoC designer. The move comes after reports that YMTC would delay its 128-layer NAND production, but industry watchers think the bankruptcy is unlikely to disrupt operations of these assets.

Over $30 Billion in Debt

Tsinghua Unigroup is owned by two major investors: Tsinghua Holdings, a wholly-owned subsidiary of the state-owned Tsinghua University, and Beijing Jiankun Investment Group, a private investment group founded by Zhao Weiguo, who happens to be the chairman of Tsinghua Unigroup. In recent years, Tsinghua Unigroup made several high-profile acquisitions funded by the government and bonds sold to various creditors, including the Hong Kong-listed state-owned Huishang Bank.  

This bank has requested the First Intermediate People's Court of Beijing Municipality to start insolvency procedures for the conglomerate since the latter had missed numerous bond payment deadlines since November, reports Nikkei Asia. As of mid-2020, Tsinghua Unigroup owed approximately $31 billion and had $8 billion in cash and cash equivalents, according to Reuters. As of mid-2020, Tsinghua Unigroup owed approximately $31 billion and had $8 billion in cash and cash equivalents. By the end of 2020, the company had defaulted or cross-defaulted on seven bonds worth $3.6 billion, the news agency says. Tsinghua Unigroup said it would cooperate with the court. 

"[We] will fully cooperate with the court to conduct a judicial review in accordance with the law, actively promote debt risk mitigation and support the court in safeguarding the lawful rights and interests of creditors," the statement by the company reads.

An Instrumental Company

Tsinghua Unigroup is instrumental for the Chinese 'Made in China 2025' initiative to build a self-sufficient semiconductor industry. The company controls some of the most successful Chinese semiconductor and high-tech companies that it has acquired over the years. 

Among these companies is Yangtze Memory Technologies Co. (YMTC) that owns a fab that can produce both 3D NAND memory and logic. YMTC can successfully compete with the world's leading 3D NAND suppliers in terms of density and performance, courtesy of its Xtacking technology. Yet, its production volume is fairly low compared to giants like Samsung. YMTC recently had to delay mass production of its 128-layer 3D QLC NAND 1.33 Tb memory devices. 

Another jewel in Tsinghua Unigroup's crown is Unisoc, a large developer of mobile phone SoCs that has consolidated Spreadtrum and RDA Microelectronics (for about $2.7 billion) and now controls a large portion of the Chinese smartphone SoC market, competing rather successfully against such companies as MediaTek and Qualcomm. Back in 2014, Intel invested up to $1.5 billion in Unisoc for a 20% stake.  

Yet another notable company in Tsinghua Unigroup's portfolio is Unisplendour, a major IT hardware, software, and services provider that carries a wide range of servers and resells Western Digital's storage systems in China. As demand for servers is on the rise, Unisplendour posted a 30% year-over-year revenue increase in Q1 2021.

Bankruptcy? Not A Problem!

Over the years, Tsinghua Unigroup has attempted to make really big overseas acquisitions to gain technologies and receive international recognition. It tried to buy Micron Technology for $23 billion and a 15% stake in Western Digital for $3.78 billion in 2015. Both deals were essentially blocked by the Committee on Foreign Investment in the U.S. (CFIUS). 

Given how important Tsinghua Unigroup is for the Chinese government and its quite competitive portfolio of companies, it is unlikely that the insolvency will lead to any disruptions of their operations.  

"Semiconductors are a priority for China's government," a source told Nikkei Asia. "Tsinghua Group itself may undergo a court-led restructuring, but any effect on how group companies are run would be limited."

ChangXin Memory (CXMT), is now officially the only Chinese DRAM maker to be in production on the market. The company has completed its Fab 1 and R&D facility located in Hefei, the capital of Anhui province, as reported by EE Times this week. 

The fab is producing 20,000 wafers per month and will be able to produce 40,000 wafers per month by the second quarter of 2020. Other Chinese DRAM makers have either shut down or are years away from production.

The CXMT facility started using a 19nm process earlier this fall to manufacture LPDDR4, DDR4 8Gbit DRAM products. CXMT built an infrastructure around its facility to house 3,000 employees, along with their families.

What About CXMT’s Local DRAM Competitors?

Not long ago, several Chinese companies attempted to design and build their own DRAM chips to compete against the likes of Samsung and Micron. It became even more imperative for the Chinese government to help create competitive DRAM makers once the trade war with the U.S. started. 

Fujian Jinhua Integrated Circuit Company (JHICC), another Chinese DRAM company, was basically killed by U.S. sanctions after the U.S. government indicted it with stealing trade secrets from Micron. JHICC shared some IP with United Microelectronics Corp. (UMC), which the U.S. Department of Justice also accused of corporate espionage against Micron.

The only other Chinese competitor besides CXMT is Tsinghua Unigroup, which is supposed to complete its DRAM fabrication facility in 2021. However, TrendForce believes that even if its DRAM products end up being competitive, it could take 3-5 years before Tsinghua can ramp up production to significant volume.

However, if CXMT does become a strong contender in the market, that could be sufficient to satisfy most of the demand in China. 

CXMT was founded in 2016 by the Hefei Industrial Investment Fund and GigaDevice and is run by GigaDevice’s former president, Yiming Zhu. The company has been recruiting engineers from Korea and Taiwan to build its fundamental DRAM knowledge, with 70% of the 3,000 employees being technical staff. CXMT also hired Karl Heinz Kuesters, former VP of technology and pre-development at memory company Qimonda.

On August 27, Tsinghua Unigroup announced that it had signed an agreement with the Chongqing government to build a DRAM fab in Chongqing’s Liangjiang New Area. The facility will begin construction this year, and it’s scheduled for completion by the end of 2021. 

TrendForce noted in its new report that the fast pace of the construction is due to the Chinese government’s commitment to becoming an independent producer of memory chips, among other things.

Long before the recent trade war with the United States began, the Chinese government already had a plan to dominate the technology sector. Some of that plan became more clear and evident when the Chinese government revealed its “Made in China 2025” policy, which aims to use government subsidies, state-owned companies, as well as acquired intellectual property, to catch up to, and then surpass western technology industries.

However, the recent trade war with the U.S. essentially gave China no choice but to bet everything on that plan and try to sustain its economy through its own technology industries. That seems to have prompted more aggressive investments in new sectors, including the memory chip production sector.

TrendForce believes that even though the fab could be completed in a short time, questions remain about how good Tsinghua’s process technology will actually be. Unlike JHICC and CXMT, two other major players in the new local Chinese DRAM market, Tsinghua doesn’t have an external partner that can provide the needed expertise. 

JHICC obtained its know-how from the Taiwanese UMC before the U.S. sanctions hit the Taiwanese companies. Similarly, CXMT obtained its expertise from Qimonda, a German memory maker that split out from Infineon Technologies back in 2006.

According to TrendForce, if Tsinghua has to set up its own process technologies, then it could take as much as 3-5 years before we see serious DRAM production from the company. 

This isn’t Tsinghua’s first attempt to make DRAM products. It first wanted to develop DRAM chips as early as 2014, but it didn’t have enough government backing at the time. At the time, it had to give up those plans and focus on NAND chip production instead.

JHICC has met with some U.S. export restrictions, and CXMT has remained the only unaffected Chinese DRAM company with plans to mass-produce DRAM chips by the end of 2020. However, the Chinese government wants more than one company to sustain the local DRAM production, which is why it has regained interest in supporting Tsinghua’s plans to develop and manufacture DRAM products.

Tsinghua and Intel unveiled more details about the "Jintide" Xeon processors here at Hot Chips 31.

Intel has a partnership in China with Tsinghua University, long known to be heavily influenced by the Chinese government, and Montage Technology Global Holdings, a Chinese server vendor, that allows the development of a co-processor that connects to Intel's Xeon cores to provide China-approved encryption/decryption technologies. These two solutions are combined in a single "Jintide" package.

Given the ongoing U.S.-China tradewar, which finds the U.S. government banning joint ventures with Chinese interests (like the banning of AMD's JV), the presentation comes at a curious time.

Tsinghua replaced the register clock driver and the data path with their design and also added a memory tracing chip and an I/O tracing chip. They integrated two extra chips into the CPU package.

Motivation: Hardware Security and Dynamic Security Check

Design flow is complex, so production of a modern chip passes through many countries. There is no way to tell if there are modifications to the chip, so China wants to monitor the chips. It is almost impossible to prove a chip is 100% secure. The only way to fix the issue is run-time surveillance to make sure the chips aren't compromised.

Three years ago the company designed a few tracking chips that can verify the behavior of the primary compute chip. They trace the system and CPU behavior and make sure that it matches expectations. This is called a dynamic security check.

Jintide Platform: Architecture and System Features

The basic platform consists of a standard server platform and they use tracking chips to monitor the CPU, memory, and all I/O coming in and out of the server. The only way to trace the hardware correctly is to trace it at the hardware level. Software won't work.

The system can identify good behavior, ignore non-harmful behaviors, yet report suspicious behaviors. Here we also see a list of the items the external chips trace. It is key to retain performance, so reducing overhead from the tracing process is a primary goal.

This is done with three total chips (fabbed on 28nm TSMC). These chips are connected through the on-die PCIe lanes on the Skylake chip. These are probably the spare lanes that Intel disables.

The MTR (memory tracing chip) tracks memory buffers, and ITR and RCP chips track the processor and other I/O. A special firmware also does checking and tracing. On the right is the check flow during run time.

The chips trace at 100ms intervals. For Skylake processors they use 52 lanes of upstream and downstream checks, needs to store 10.4 MB.

Jintide Chips: Specification and Tapeout Results

This chip is the I/O tracing chip. 300mm squared running at a 40W TDP. That's quite a bit of power for surveillance.

The second chip is the RCP. It has three microprocessors inside, and there are encryption and decryption features in the chip. There are also two reconfigurable arrays.

Here we can see the multi-chip package they designed. The pin-out is almost the same as a regular Skylake, but it has two extra chips, so power consumption is higher.

The MCP also serves as a root of trust for the system. It also detects and identifies all of the devices connected to the server.

The system also offers encryption features through the Jintide Open API that runs in a custom hypervisor that also needs installed in the system. The system also generates and manages the platform's encryption keys.

The average performance loss is less than 10% due to the surveillance on the server. It's noteworthy that these servers are only sold in China.

Here we see the performance impact of different polling intervals.

These are an examples of the system detecting a trojan, and another example of detecting vulnerabilities. Here they use Spectre as an example. The company believes that by extending their model they could even find microarchitectural vulnerabilities.

Jintide: Conclusion

The Jintide processors are largely designed to trace the behavior of the server to assure that they aren't compromised, and the custom chips the company designed can do so with a minimum of performance loss.

It's noteworthy that the chips have unfettered access to all system memory, I/O, and the CPU, not to mention control the encryption key generation and management. That means if the control chips were compromised, an attacker would have unfettered transparent access to the entire system. As such, the company also has encryption features in their third-party chips.

China’s Tsinghua Unigroup announced today that it's entering the DRAM market, making it the third Chinese company to do so in recent years. 

Tsinghua Unigroup announced that it formed a new business unit for producing DRAM in a one-sentence statement released on Sunday. According to analyst Trendforce, even though Tsinghua has only just started working on its DRAM division, which it described as being in "early stages," it shouldn't have a problem completing the division's development. Tsinghua first talked about building a DRAM division in 2014, but those plans were put on hold in favor of NAND flash design, Trendforce noted. 

The DRAM unit is funded mostly by Chinese government money, just like the DRAM businesses of two other Chinese companies: Jinhua Integrated Circuit Company (JHICC) and Changxin Memory Technologies (CMT, previously called Innotron Memory).

In November, JHICC was indicted by the U.S. government for stealing DRAM technology from Micron, a U.S. supplier of memory technologies. Soon after the indiction, CMT also changed the designs of its memory technologies to avoid any potential infringement on U.S. companies’ patents.

Trendforce analysts believe Tsinghua’s plans were revived due to JHICC being put on the U.S. Entity List last year, banning U.S. companies from supplying it with tech. The ban almost forced JHICC to go out of business, and Trendorce analysts don’t believe that CMT’s memory division will be enough to satisfy the local demand or Chinese government hopes for independence from U.S. companies.

Intel will end its 5G technology sharing partnership with Unisoc, China’s second largest mobile chip maker. The venture was announced less than 12 months earlier, acording to a report today by the Nikkei Asian Review.

Last year, Intel announced a “multi-year” venture that would give it access to China’s market. In return, Intel would share all of its 5G technology with Unisoc, the maker of Spreadtrum chips. Short of 12 months later, the partnership has ended. 

In 2014, Intel invested $1.5 billion for a 20 percent stake in China’s Tsinghua Unigroup, which owns Unisoc (formerly known as Spreadtrum). That partnership was based upon Spreadtrum building systems-on-chip using Intel’s Atom microarchitecture, but that partnership also came to an end last year, when Spreadtrum announced it would stop making Atom-based chips. The announcement surprised no one considering Intel itself had previously announced it would exit the mobile SoC market.

The U.S. government has recently barred federal agencies from using devices or equipment made by some Chinese firms. Furthermore, the U.S. government has also launched a global campaign against Huawei, warning that the Chinese company’s equipment may contain backdoors that would put critical infrastructure at risk. The U.S. government has also asked allies to stop sharing intelligence information over Huawei-provided equipment, if the allies continue to use Huawei gear.

Robert Topom, GM of Intel's 5G Strategy and Program Office, told Nikkei Asian Review that the decision was "just made recently." The publication noted Intel denied that the partnership was halted due to pressure from the U.S. government.

Yet, an unnamed source Nikkei Asian Review claims has "direct information" said that while former Intel CEO Brian Krazinch was a favor of the deal, "Intel has been concerned the closer ties with the Chinese government-sponsored company could somehow upset U.S. authorities given the current tensions between the two sides." Unisoc declined to comment to the publication.

Unisoc also announced that it plans to release a 5G modem that was built in-house without any help from Intel. Smartphones using Intel’s own 5G modems are expected to ship in 2020, and Apple is expected to be one of the main customers after its recent spat with Qualcomm.

Tsinghua Unigroup, one of the many tentacles of the Chinese state-controlled Tsinghua University, announced that it is investing $30 billion to build a new DRAM and NAND fab in Nanjing, China. The development comes as the company seeks to rapidly expand after its 51% buy-in of Yangtze River Storage Technology, which recently announced a separate $24 billion DRAM and NAND fab in Wuhan, China.

The NAND industry is falling deeper into the largest shortage in its history, and many industry analysts predict that we will experience yet another DRAM shortage in the coming months. Both the NAND and DRAM industries have consolidated down to a few key players, and predictable production output has kept supply and demand dynamics largely balanced for several years. However, a rash of 3D NAND developmental delays have delayed major players, such as SanDisk, Toshiba, SK hynix, and to a lesser extent Intel and Micron, from reaching production projections. These delays are the catalyst for the current shortage, but the entrance of the potentially unpredictable Tsinghua could upset the delicate supply balance, thus causing an eventual glut. The staid semiconductor industry would rather weather shortages than the margin-killing gluts, so the Tsinghua developments are concerning for the established players.

The Chinese government is diving into the semiconductor business with aplomb, as it cites the lack of native production as a national security risk. The brass ring of flash IP has eluded Tsinghua as it attempts to buy its way into the semiconductor business. The first of its fruitless efforts came in the form of a $2.8 billion investment in Western Digital (through its Unisplendour subsidiary), which at the time was in the midst of purchasing SanDisk for $19 billion. The strategic investment promised to grant Tsinghua access to the patent-protected IP it needs to jump-start production, but it later withdrew the offer after the US Committee on Foreign Investment in the United States (CFIUS) threatened to scuttle the deal based on the US' own national security interests. Undeterred, Tsinghua also reportedly began talks with both Micron and SK hynix, but neither initiative came to fruition.

With those avenues closed, the company proceeded to deepen its investment in home-grown projects and hired several high-level Taiwanese semiconductor executives to bolster its capabilities. The company projects that its new Nanjing, China facility will produce up to 100,000 wafers per month and will begin pilot production in 2018. The company will likely need to license several key aspects of the technology, and it's unclear if it will select floating gate or charge trap technology, though the latter is more likely. Tsinghua might attempt to invest in Toshiba's pending semiconductor spinoff. WD appears to be the key player in the Toshiba negotiations thus far, but if a determined Tsinghua approaches, it's possible it could offer Toshiba an enticing offer.

Other experienced and established industry stalwarts have faced significant challenges while transitioning to smaller DRAM lithographies and 3D NAND, so the multibillion-dollar question is if the fledgling Tsinghua initiatives can jump the hurdles in a time-effective and economical manner. However, if the Chinese government-backed initiatives succeed there is no question that it will drastically alter the status quo. The country also has its sights aimed squarely at producing its own processors, so its investments in key semiconductor technologies is far from over.

Magnetic memory devices that are controlled by voltage and not by heat have been researched for decades, but research earlier this year showed that the coercive force of magnetic materials is controlled by heat and not by voltage.

The Chinese scientists now claim that those research results were wrong and that coercivity can be controlled by voltage. Coercivity is the intensity of a magnetic field that is necessary to demagnetize a magnet when it is fully magnetized.

The hope of the storage industry is that using voltage-induced magnet control will lead the path to storage technologies that use much less power than the products we know today.

The scientists said that they explored coercivities in three structures that are "commonly used in magnetic memory experiments" and discovered that heat is not at all responsible for the changes of a magnet's coercivity, but voltage is. Without providing details publicly, the Journal of Applied Physics, which published their paper, said that their results show that voltage is directly controlling changes in the magnetic properties of all three of the tested materials. "For example, the researchers demonstrate that the effect can be turned on and off almost instantaneously, whereas the changes should lag if heat is the cause."

There was no information what further research is necessary and how quickly the findings could result in actual new technologies and new products.