Charts
DataOn-chain
VIP
Market Cap
API
Rankings
CoinOSNew
CoinClaw🦞
Language
  • 简体中文
  • 繁体中文
  • English
Leader in global market data applications, committed to providing valuable information more efficiently.

Features

  • Real-time Data
  • Special Features
  • AI Grid

Services

  • News
  • Open Data(API)
  • Institutional Services

Downloads

  • Desktop
  • Android
  • iOS

Contact Us

  • Chat Room
  • Business Email
  • Official Email
  • Official Verification

Join Community

  • Telegram
  • Twitter
  • Discord

© Copyright 2013-2026. All rights reserved.

简体繁體English
|Legacy
BTCBTC
💲79876.76
-
1.33%
ETHETH
💲2268.73
-
1.4%
SOLSOL
💲91.02
-
4.4%
DOGEDOGE
💲0.1142
+
1.6%
USDCUSDC
💲1.00
-
0%
XRPXRP
💲1.43
-
2.05%

qinbafrank
qinbafrank|5月 12, 2026 12:20
Can China and the United States reach a deal for rare earth exchange lithography machines? If it is achieved, will storage immediately become less scarce? How do you see the turning point of storage in the future? Today's Barron Weekly article has spread widely. The core argument is that China and the United States may conclude a rare earth lithography machine transaction during Trump's visit to China, after which TSMC and Samsung Hynix will lose their chip pricing power. Is that really the case? There are several issues and misconceptions that need to be discussed here 1. What is the probability of China and the United States reaching a rare earth exchange for lithography machines? I also talked about Trump's visit to China yesterday in a forward-looking tweet that compared with Nvidia Chip China, China wants more semiconductor devices such as lithographers on key choke points, which is earlier than what Barron Weekly talked about. Yesterday we also talked about how China is not very interested in lifting the ban on Nvidia chips, but rather wants to support domestic chips. But key semiconductor equipment (DUV, EUV lithography machines) and high-end EDA software are very scarce, and China should want these devices. But can China and the United States reach such an agreement? I am not sure about this because even though China implemented rare earth control measures in October last year, the United States did not loosen its grip on critical semiconductor equipment. Especially in the past 26 years, the United States has further tightened its restrictions on China's semiconductor critical equipment. I can only say that it depends on the outcome of the conversation 2. If China and the United States really reach a deal for rare earth exchange lithography machines, can domestic storage immediately increase? Actually, it's not like that China has indeed been stuck in advanced processes, especially in EUV/DUV lithography, for many years. After obtaining ASML equipment, the acceleration of localization is a high probability event. Chinese memory manufacturers are mainly domestic DRAM players such as CXMT Changxin Storage. In the past two years, they have made rapid progress in mid to low end memory and have also seized a considerable share. There is still a significant gap between the previous generations of mid to high end memory. But it should also be noted that the lithography machine is indeed one of the most critical bottlenecks at present, but it is far from the only reason for the large generation gap. 1) The lithography machine (DUV/EUV) is indeed the biggest drawback EUV (Extreme Ultraviolet Photolithography) is currently completely unavailable in China, and Samsung, SK Hynix, and Micron have used EUV in advanced DRAM (HBM4) and some high-density NAND to significantly improve density and yield. Chinese manufacturers (CXMT, YMTC) can only rely on DUV+multiple patterning (SAQP, etc.) for rigidity, with complex processes, high costs, and low yield rates. 2) Advanced DUVs (such as ASML NXT: 2000i/2050i series): Currently, some can still be imported, but new laws such as the US MATCH Act in 2026 have directly blocked high-end submerged DUVs and their maintenance/parts for CXMT, YMTC, etc. China is accelerating the use of domestically produced SMEE SSA800 and other devices to replace them, but these devices are currently mainly suitable for 28nm and above, and achieving advanced nodes still requires multiple exposures, with efficiency far lower than ASML's latest models. The direct consequences of these two are: CXMT's advanced DRAM yield is about 50% (leader 80-90%), while YMTC's high-level NAND is in mass production, but its cost and stability are limited. This directly leads to weak competitiveness of mid to high end products (DDR5/LPDDR5X, HBM3, 300+layer NAND). If we only look at this one point, lithography limitations can indeed explain 60% of the current gap - without it, Chinese manufacturers simply cannot efficiently climb advanced nodes. But there are other core gaps that cannot be solved instantly by unsealing the lithography machine 1) Patent/IP barriers: Samsung, SK Hynix, Micron have a large number of patents on DRAM/NAND core architecture (Xtacking, HBM stacking, etc.). Chinese manufacturers either take detours (due to high costs) or are restricted from directly taking the optimal path. 2) Incomplete equipment ecosystem: In addition to ASML lithography, advanced models of American and Japanese equipment such as Lam Research, Applied Materials, and Tokyo Electronics are also limited. 3D NAND vertical stacking and HBM hybrid bonding require a complete set of high-end toolchains. 3) Yield and process integration: Even if the equipment is in place, process debugging, defect control, and material matching all require time. The slow ramp up of yield for CXMT/YMTC new lines is a recognized industry issue, which is not only related to lithography machines, but also involves the coordination of the entire process such as etching, deposition, CMP, etc. Especially the debugging procedures for extreme ultraviolet lithography machine EUV are more complex, with longer time and process. So not only is there a lack of high-end lithography machines, but also the gap between domestic memory manufacturers and Samsung, Hynix, Micron is created by patents, equipment ecology, process integration, and other factors. Lithography is a bottleneck, but there are still several hurdles to consider, including IP/patents, ecology, and craftsmanship. There is actually a time difference, even if an agreement is signed, ASML equipment delivery, production line construction, and yield ramp up will take at least 1-2 years to form substantial production capacity. From this perspective, even if Chinese manufacturers obtain the most advanced semiconductor equipment and it takes basically 28 years to form production capacity for mid to high end storage, they still overcome obstacles such as IP/patents, ecology, and processes, and everything goes smoothly. So we can't ignore Samsung, Hynix, and Micron's own expansion plans. In the past six months, three storage companies have also started actively expanding production, including Samsung P4/P5 factory and Hynix Yongin factory? New lines such as Micron Tongluo/ID1 will only form substantial growth in the second half of 2027 to the first half of 2028 at the earliest. This means that the high-end production capacity of domestic storage manufacturers has not yet been formed, and the expansion production lines of Samsung, Hynix, and Micron have begun to release production capacity. 3. So what do you think about the turning point of storage in the future? My personal framework is based on demand and supply: 1) When will demand slow down? The current situation is that with the rapid iteration of AI and the rapid popularization of long context, multimodal, and agent technologies, the consumption of storage is very high, and demand is still growing rapidly without any signs of slowing down. 2) When will the supply be increased? In fact, the second part just talked about increasing supply volume. Now it is clear that the new production lines of Samsung, Hynix, and Micron have also released their production capacity by the end of the 27th. If China and the United States can reach an agreement on rare earth exchange for lithography machines, and Chinese storage manufacturers obtain the most advanced lithography machines, it will take 28 years for production capacity to land smoothly. So at least in the first half of next year, the production capacity of storage (especially high-end HBM, GDDR7/GDDR6x, DDR5, mid-range DDR4, high-end NAND) will still be scarce, depending on the sustained high-speed growth of demand. And the demand is still skyrocketing right now The turning point for future storage will either be a slowdown in demand or an increase in supply. Just keep an eye on these two points
+5
Mentioned
|
APP
Windows
Mac
Share To

X

Telegram

Facebook

Reddit

CopyLink

|
APP
Windows
Mac
Share To

X

Telegram

Facebook

Reddit

CopyLink

Timeline

5月 14, 02:07Binance Alpha 2.0 supports Baby Shark brand upgrade
5月 13, 19:13Remove support for six non-USD trading pair assets
5月 13, 12:25Coinbase drives revenue diversification through stablecoins and derivatives business
5月 13, 09:46Bithumb will list the UP token on the Korean Won market
5月 13, 08:54Vietnam plans to launch a regulated cryptocurrency exchange in 2026
5月 13, 04:05Crazy Wednesday Digital Storage Special Trading Event
5月 13, 03:35HERMESOS/USDT trading pair officially launched on WEEX
5月 12, 13:55People Dex supports RWA perpetual contracts
5月 12, 12:00HTX launches double rewards for new users' first spot trading
5月 12, 10:40100 Product Requirements Eliminated by Agent Trade Kit

HotFlash

|
APP
Windows
Mac
Share To

X

Telegram

Facebook

Reddit

CopyLink

APP
Windows
Mac

X

Telegram

Facebook

Reddit

CopyLink

Hot Reads