memo

According to the Financial Times, citing informed sources, Chinese memory chip manufacturers Changxin Storage and Yangtze Memory Technologies have purchased enough deep ultraviolet lithography machines (DUV) from ASML to meet their expansion plans for the next three years.

KB Securities warned that the inventory of memory semiconductors at Samsung Electronics and SK Hynix has fallen to less than 10 days of supply, and the available supply next year will be significantly insufficient. The agency reported on Monday that investment in artificial intelligence infrastructure is expanding at an unprecedented pace, leading to a severe supply shortage in the memory chip market. A key factor is the transition to HBM4, which requires about three times the wafers of traditional DRAM, and limited wafer capacity will reduce the available capacity of traditional DRAM.KB Securities expects that next year, the demand for DRAM and NAND will exceed supply by more than 10 percentage points. Research director Kim Dong-won stated that artificial intelligence servers will consume HBM, server DDR5, and enterprise-grade solid-state drives, potentially leading to a historic shortage. Global hyperscale data center operators have raised their expectations for artificial intelligence infrastructure investment next year to $1.3 trillion, a 60% increase from the previous year. The agency anticipates that memory semiconductors will account for 57% of total investment in artificial intelligence infrastructure next year, up from 14% last year, with TrendForce estimating this ratio could reach as high as 68%. Samsung Electronics' stock price has dropped nearly 28% from its peak, while SK Hynix has fallen nearly 40%.

According to Blue Whale News, from Changxin Technology, ByteDance and ZTE have jointly developed a new generation of Doubao phone------Nubia NaviX Ultra, which will be equipped with Changxin's highest speed LPDDR5X memory at 10667Mbps. The official statement indicates that this is also the first time that domestically produced 10667Mbps high-speed LPDDR5X memory chips have achieved mass production and application.Previous news reported that the Nubia "Doubao phone" has completed network filing and will be mass-produced and launched in September.

The IPO review status of Yangtze Memory Technologies Co., Ltd. for its initial public offering and listing on the Sci-Tech Innovation Board has been changed to "Inquired.""Inquired" means that Yangtze Memory has officially entered the listing review stage at the Shanghai Stock Exchange, where a round or multiple rounds of written Q&A will be conducted regarding the company's information disclosure, Sci-Tech attributes, and other issues.Yangtze Memory is China's largest 3D NAND flash memory manufacturer and the only 3D NAND original factory. It has independently developed the Xtacking architecture and is a memory IDM company that integrates chip design, production manufacturing, and packaging testing.

According to Cointelegraph, the Dubai Virtual Assets Regulatory Authority (VARA) has signed a Memorandum of Understanding (MoU) with the BlackRock-backed tokenization platform Securitize to advance the tokenization and digital asset infrastructure development in the UAE and Dubai. Both parties stated that the agreement will establish a collaborative framework to support regulated tokenization initiatives, promote institutional participation, and strengthen Dubai's digital asset ecosystem.A VARA spokesperson stated that the memorandum aims to combine VARA's regulatory perspective with Securitize's experience in institutional tokenization to determine how the collaboration can help Dubai develop a trustworthy and regulated tokenization market, with no specific projects to be announced at this stage. Carlos Domingo, co-founder and CEO of Securitize, stated that Dubai is one of the most forward-looking jurisdictions for global digital asset innovation, and collaborating with regulators is crucial for tokenization to move from concept to mainstream financial infrastructure.At the time of this announcement, investor demand for tokenized assets continues to rise. According to RWA.xyz data, the total number of tokenized asset holders has increased by 103% over the past 30 days to 3.2 million, and the total value of tokenized assets has grown by 2% to $38.5 billion. Securitize is currently the largest tokenization platform globally, with a tokenized asset management scale of $4.9 billion, while Ondo Finance ranks second with $3.5 billion.

Google's Senior Director of Supply Chain Infrastructure, Nikhil Cherian, revealed that to overcome memory bottlenecks, Google is developing software and hardware solutions and dismantling retired servers to recycle DDR4 components to establish an internal recycling supply chain. Google has designed special hardware adapters to connect the previous generation DDR4 to the new generation of AI servers while importing retired servers to remove their DDR4 modules for recycling.Cherian stated that the AI industry has rapidly shifted from being compute-constrained to memory-constrained, with high-performance memory accounting for about 75% of the bill of materials cost for a given AI server. A Goldman Sachs report indicated that memory prices will continue to rise in the third quarter, with personal computer DRAM prices expected to increase by 18% to 23% and server DRAM prices expected to rise by 13% to 18%. Trendforce data shows that in August, the spot market prices for DDR4 8GB and DDR5 8GB rose to $142 and $133, respectively.The two TPU ASICs launched by Google this year have been optimized for memory design, claiming to reduce memory demand to one-sixth of the original. The TPU8i chip features a dedicated layered memory design that relies on a high-speed DDR5 memory architecture to perform host-level tasks, with each chip equipped with 288GB of HBM3e high-bandwidth memory.

The SEMICON Taiwan 2026 Memory Summit took place on the 1st, where Nikhil Cherian, Senior Director of Supply Chain Infrastructure at Google Cloud under Alphabet, pointed out that with the popularity of multimodal and mixed expert architectures, AI computation has shifted from being power-limited to memory-limited, with high-performance memory accounting for over 75% of the cost of AI server hardware bill of materials. In the face of capacity, bandwidth, and power consumption bottlenecks, Google is breaking through the AI memory bottleneck through a dual-track strategy of hardware offloading for inference and training, and lossless quantization software algorithms.Google adopts an offloading strategy in hardware architecture, launching TPU 8i for low-latency inference and TPU 8t specialized for large-scale training. The TPU 8i is equipped with 288 GB of high-bandwidth memory, with SRAM capacity on the chip increased threefold to 384 MiB, placing dynamic conversation states and key-value caches on the chip itself to achieve zero chip-off latency. The TPU 8t forms a super-large computing cluster with 9600 chips, achieving a shared pool of HBM at a scale of 2 PB, eliminating chip-off data transfer bottlenecks, along with TPU Direct Storage technology.Google has developed the training-free TurboQuant lossless quantization algorithm, compressing the key-value cache of large models from 32 bits to 3 bits, reducing memory usage by six times without loss of accuracy, resulting in an eightfold acceleration in attention computation, and integrating old-generation DRAM technology to extend the lifecycle of components.

Google announced two new quality requirements for Android applications to address the industry memory chip shortage triggered by the AI data center boom. One focus is on reducing in-app memory usage and optimizing code. Google stated that the mobile industry is facing significant hardware supply constraints, which are changing the availability of device memory and, in turn, affecting user experience.To this end, Google has set new performance thresholds in areas such as dynamic memory usage and bitmap usage, and has increased code optimization requirements to prevent app stuttering and crashes. The company will launch new tools to alert developers to optimize when applications exceed the thresholds. More diagnostic tools will be released later this year, including a Memory Limiter feature that prevents applications from excessively occupying device memory.Developers must meet the new thresholds by February 2027. Additionally, all Play Store applications must comply with the Zero Tap Sign-In requirement by April 2027, which automatically restores user login status during device migration via the Android Restore Credentials API. These changes aim to help app and game developers continue to provide high-quality experiences in a market with limited memory supply, especially targeting price-sensitive low-end devices.

According to Nikkei Asia, SK Hynix CEO Kwak Noh-Jung stated that the global shortage of memory chips is expected to last until the end of 2030, with a low risk of oversupply. He pointed out that in the AI era, memory chips are no longer just "commodities," and there are currently no signs of oversupply or industry downturn; demand from AI customers remains strong.SK Hynix is a major supplier to NVIDIA. The report stated that the company remains open to further expanding its investments in the U.S. after launching its first HBM chip packaging plant in Indiana.

Samsung Electronics and SK Hynix plan to increase the supply of 8-layer HBM4 memory to NVIDIA in the second half of this year.This move is influenced by NVIDIA's supply strategy, aiming to ensure a stable supply of HBM memory while also considering product heat generation.

According to South Korea's Chosun Ilbo, the export prices of DRAM in South Korea continue to soar, with AI-driven HBM production squeezing the supply of ordinary memory. Data shows that from the 1st to the 20th of this month, the export unit price of DRAM from South Korea reached $92,183 per kilogram, a year-on-year increase of 401%, and approximately 12.5 times higher than the low point in January 2023.Goldman Sachs expects the DRAM supply gap to widen from 5.0% this year to 5.9% next year, believing that the HBM required for AI servers and high-capacity server DRAM are absorbing limited production capacity, leading to further tightening of ordinary DRAM supply, with shortages potentially lasting until 2028. TrendForce predicts that by the end of next year, the wafer input for HBM production by Samsung, SK Hynix, and Micron will account for 30% of the total DRAM wafer input, but HBM will only account for about 13% of the actual DRAM bit supply.Memory manufacturers warn that the supply tightness of DRAM and NAND driven by AI demand may continue until after 2027. SK Hynix stated that from the supply side, 2027 could become the most severe shortage year in the history of the memory industry.

According to DIGITIMES, Changxin Memory and Yangtze Memory continue to expand, and recently Apple is actively seeking to introduce the two Chinese memory supply chains, drawing attention to their entry into Apple's supply system. Industry insiders reveal that the two manufacturers are actually using Apple to prove that the quality of Chinese DRAM and NAND Flash has reached international standards, while secretly expanding their cloud and enterprise market share through third-party assembly for export to North American neocloud operators.Changxin Memory's parent company, Changxin Technology, has completed fundraising of 29.5 billion yuan, while Yangtze Memory plans to raise about 33 billion yuan through an IPO for production line upgrades and R&D. Yangtze Memory has turned profitable in 2024, with a net profit attributable to the parent company reaching 33.379 billion yuan in the first quarter of 2026; its third-phase new factory will begin production ahead of schedule in the fourth quarter of 2026, with a total monthly production capacity of 100,000 wafers, of which about 20% will be trial-produced LPDDR, becoming its first DRAM base. The supply chain indicates that the two companies are expected to increase the self-sufficiency rate of domestic memory in China to over 50% by the end of 2027 to 2028.Yangtze Memory sells NAND to third-party module manufacturers and assembles them into enterprise-level SSDs to avoid location sensitivity; Changxin Memory has already obtained certifications from small and medium-sized cloud operators in the U.S., Canada, and other regions. U.S. private enterprises can, in principle, procure from the two companies, with the entity list primarily restricting their access to U.S. technology, but not completely prohibiting private procurement. Cloud operators also reduce compliance risks through methods such as renting computing power or holding assets through third parties. The industry believes that Apple's influence on supply and demand in the industry is declining, and the two companies have no intention of supplying Apple in large quantities or engaging in price competition, making it difficult to meet Apple's demand in the Chinese market in the short term.

Xiaomi officially released the Xuanjie O3 chip. The new generation Xuanjie O3 is the world's first flagship mobile processor (SoC) that supports LPDDR6 memory chips. According to reporters from the Financial Associated Press, domestic storage leader Changxin Storage (CXMT) is a core partner for Xiaomi's Xuanjie O3 LPDDR6 memory.In March this year, there were reports that Changxin LPDDR6 had sent samples to key customers, and in early August, multiple media outlets reported that it was nearing the end of verification. With the launch of the Xuanjie O3, this marks the first application of domestic LPDDR6 chip products in flagship processors, which also means that domestic low-power memory has officially entered the global top tier.

According to the Chosun Ilbo, in the context of the U.S. government increasing pressure on investment in South Korea, the U.S. has recently made demands to the South Korean government and companies to build storage semiconductor production facilities in the U.S. and ensure a stable supply of storage chips. In response to South Korea's promotion of a 800 trillion won semiconductor cluster plan, the U.S. has also informally expressed dissatisfaction, believing that South Korea is actively supporting domestic semiconductor investments while showing a negative attitude towards investing in the U.S. This is the first time the U.S. has mentioned the South Korean government's "super project" while pressuring for investment.Minister of Trade, Industry and Energy Lee Chang-yang also discussed related matters during his recent visit to the U.S. for investment negotiations. The South Korean government believes that the U.S. semiconductor investment demands are independent of the previously agreed $350 billion investment arrangement with the U.S., and plans to announce its first investment project in the U.S. in September. Domestic semiconductor companies are concerned that if they formally accept the U.S. request to establish factories while promoting the Hunan cluster, it may create a dual burden. Additionally, it is reported that Foreign Minister Park Jin spoke with U.S. Secretary of State Rubio that evening to exchange views on investments in the U.S. and bilateral relations.

On August 24, according to Reuters, China's largest flash memory chip manufacturer Yangtze Memory Technologies Co., Ltd. (YMTC) and its parent company Yangtze Memory Technologies Holdings (CCSH Corporation) have had their listing application accepted by the Shanghai Stock Exchange's Science and Technology Innovation Board, with the market expecting the company to go public next year. This IPO is expected to raise 33 billion yuan (approximately 4.9 billion USD). CCSH plans to issue between 1.98 billion and 2.43 billion new shares.The prospectus shows that the company's revenue in the first quarter of 2026 (January to March) reached 47 billion yuan, nearly a fivefold increase year-on-year; net profit was 33.38 billion yuan, more than double the net profit of 14.21 billion yuan for the entire year of 2025. YMTC's factories are currently operating at near full capacity and are continuously expanding sales in the tightening NAND flash memory market. The average selling price of NAND flash memory in the first quarter of 2026 reached 2.73 times the average level of 2025, and the gross profit margin also significantly increased from 35.3% in 2025 to 76.8%.CCSH plans to use 20.8 billion yuan of the funds raised from this listing to expand production lines, and an additional 12.2 billion yuan for research and development of the next generation of NAND flash memory and high-speed storage products.

The Shanghai Stock Exchange has changed the IPO review status of Yangtze Memory Technologies Co., Ltd. to accepted for its initial public offering and listing on the Sci-Tech Innovation Board, with a proposed financing amount of 33 billion yuan. The sponsoring institutions are CITIC Securities and CITIC Construction Investment.The prospectus of Yangtze Memory shows that after turning a profit in 2024, the profit level will rise rapidly, with operating revenue of 47.042 billion yuan and net profit attributable to the parent company of 33.379 billion yuan achieved from January to March 2026. According to TrendForce data, from January to March 2026, the company ranked third globally and first in China among global NAND Flash manufacturers by both sales revenue and shipment volume.

On August 21, 3D storage-computing integrated chip company Qianhe Yibang announced that its self-developed global first 4-layer 3D DRAM stacked storage-computing integrated chip successfully returned to the wafer and was powered on, marking the transition of this technology from validation to the engineering implementation stage. This chip adopts a three-dimensional integrated native computing architecture, with data access and storage bandwidth improved by an order of magnitude compared to traditional solutions, while access and storage power consumption, latency, and single data processing costs have all decreased by an order of magnitude.Qianhe Yibang was established in January 2024, incubated by NetEase, and recently completed over 2 billion yuan in Series B financing.

According to BeInCrypto, Micron CEO Sanjay Mehrotra stated at a semiconductor manufacturing plant near the company's headquarters in Boise, Idaho, that memory has become a strategic infrastructure in the AI era, and the growth in AI-driven demand will permanently reset the economic logic of this cyclical industry. In terms of financial data, Micron's revenue for the third quarter reached $41.46 billion, a significant increase from $9.3 billion last year, with a gross margin rising to 84.6%, and it is expected that the gross margin for the fourth quarter will further increase to about 86%.On the supply side, the demand from data center customers is about 150% of the supply that Micron can commit to, indicating a significant gap. Mehrotra also listed autonomous vehicles, robotics, and AI consumer devices as sources of incremental demand in the future.