供应链分析
产业链结构、上下游与瓶颈环节研究 · 共 1315 条 · 滚到底自动加载更多
-
LLM难析二阶效应，中国管制致住友短缺影响LITE，涨价或抵消。
大语言模型(LLMs)在分析二阶和三阶效应方面表现糟糕。这基本上源于中国对日本实施出口管制后的供应链冲击。住友电气(Sumitomo) -> $LITE -> $GOOGL（但如今住友可能面临短缺，因此提供给LITE的材料减少 -> 收入减少）。但如果光学组件制造商决定像高带宽内存(HBM)那样提高价格，这一影响将被抵消。
英文原文
LLMs are terrible at analyzing second-third order effects. Basically it stems from supply chain shock after China's export controls on Japan. Sumitomo -> $LITE -> $GOOGL (but now Sumitomo might face shortages, so less material supply for Lite -> less revenue). But this would be offset if optical component makers decided to ramp up prices like HBM.
-
中国出口限制加剧InP供应瓶颈，利好AXTI垄断地位。
市场正在实时见证磷化铟(InP)的供应冲击。 本周时间线如下： Fastmarket Research, 2025： 美国$AXTI和日本住友控制着近80%的InP市场。 中国商务部 - 2026年1月6日：中国宣布对日本新增“两用”物品出口限制。 日本经济新闻(Jiji) - 2026年1月9日： 中国供应商开始通知日本客户：“我们不再向日本出口稀土金属（日本定义的包含镓、铟、锗等31种矿物的组别）。” 刘军红 CGTN - 2026年1月10日： “在战略材料方面……日本对中国的依赖度相对较高。” 上海有色网(SMM) - 2026年1月12日： 7N铟在SMM创下历史新高，达到812.45 ~ 901.04美元/公斤。 评论： $AXTI本周因面临财报前的出口管制问题而股价下跌。 然而，这被广泛误解，因为这对AXT来说是一把双刃剑，最终有利于AXT。 近期的出口问题从短期收入损失（无法向日本出口）的角度，巩固了AXT在中期内对InP衬底市场的控制力。 出口管制禁令伤害了其主要的日本竞争对手，住友和JX受影响最大，因为它们可能面临原材料短缺的上限。 如果住友在几个月内耗尽原料，西方AI客户（英伟达、博通、Lumentum等）除了从AXTI购买外别无选择，无论价格如何或许可证延迟多久。 随着公司争相建立长期库存，我们已经看到原料价格实时飙升。 AXTI可能成为大规模InP晶圆的主要来源，市场正在实时关注这一瓶颈的展开。
英文原文
Markets are witnessing a supply shock happen real time with InP. Here's the timeline this week: Fastmarket Research, 2025: US-based $AXTI and Japan's Sumitomo control nearly 80% of the InP market. China Commerce Ministry - Jan 6 2026: China announces new export restrictions on "dual-use" items to Japan. Jiji via Nippon - Jan 9, 2026: Chinese suppliers have started informing Japanese clients: "We can no longer export Rare Metals (Japanese group of 31 minerals like Gallium, Indium, Germanium) to Japan." Liu Junhong CGTN - Jan 10th, 2026: "When it comes to strategic materials... Japan's dependence on China is relatively high" SMM - Jan 12th, 2026: 7N Indium hits record high on SMM at 812.45 ~ 901.04 USD/KG. Commentary: $AXTI recently dropped on facing export control issues on pre-earnings this week. However, this is widely misinterpreted, since this a double edged sword that benefits AXT. Recent export issues cements AXT's control over the InP substrate market in the medium term from short-term revenue loss (from the inability to export to Japan). Export control bans hurt their main Japanese competitors in Sumitomo and JX are impacted the most, as they may face a ceiling in from material shortages. If Sumitomo runs out of feedstock in a few months, Western AI customers (Nvidia, Broadcom, Lumentum, etc.) will have zero alternative but to buy from AXTI, regardless of the price or permit delays. We're already seeing feedstock prices explode real-time as companies scramble for long term inventory. AXTI may become the main source for InP wafers at scale and markets are watching this bottleneck unfold real-time.
-
AI模型数据污染需VLN人工审核
@ENTJ_46 例如，对于 Opus 模型，^ 它们都使用了受污染的数据，这需要借助 $VLN 进行人工审核 https://t.co/wTpg2BTVYn
英文原文
@ENTJ_46 Eg. With opus, ^ they all use contaminated data, and this requires manual review with $VLN https://t.co/wTpg2BTVYn
-
询问博通v8供应链及BOM信息，认为利好该股。
@jukan05 哇，这对 $AVGO 来说立刻就是看多的。你碰巧有 v8 供应链/物料清单(BOM) 的信息吗？
英文原文
@jukan05 Wow that’s immediately bullish for $AVGO. Do you have the info of the v8 supply chain/bom by any chance?
-
日本InP出口管制或致光模块产能瓶颈，作者减持相关风险敞口。
我们正目睹市场的第三阶效应，$LITE、$AAOI 及光子学板块遭遇血洗，跌幅超 10%。 原因多重，包括 SanDisk 下跌 10%+ 引发的板块拖累及其他抛售。 但我的细微观点是：由于日本出口管制，整个日本供应链的磷化铟(InP)库存/产能可能耗尽，而这正是西方大规模光模块建设所必需的。 我不会将 $COHR 与其他光子学玩家混为一谈，但他们已面临产能紧张，且受住友、JX 影响，情况可能恶化。 这将导致 $LITE 等缺乏 InP 来为 $GOOGL TPU Ironwood 等生产光模块的公司产能受限。 然而，若出现以下情况，这种影响将被抵消，我们可能看到持续反弹： -> 下游如高带宽内存(HBM)的定价溢价抵消产品短缺。 -> 日本以某种方式解除 InP 出口禁令（不太可能）。 -> 若 $AXTI 在获得 1 亿美元融资翻倍产能后，能向西方供应商大规模交付衬底，事情将顺利进行。 总之，我想从供应链角度提供另一种视角，因为鲜有人评论这一点。 鉴于本周看到的出口管制风险及 $AXTI 的新垄断瓶颈，我个人已移除部分光模块敞口风险。但我不是空头，不会利用这一点。 我对整体板块极度看好，但对光子学玩家（是风险厌恶抛售延续还是反弹）的走势毫无头绪。
英文原文
We’re seeing third order effects in markets with $LITE, $AAOI, and photonics having a bloodbath, dropping 10%+. It’s multifaceted, with sector drop from Sandisk -10%+ and others selling off as well. But my nuance take is that the entire Japanese supply chain will likely run out of InP inventory/capacity from Japanese export controls, required for the western optical buildout at scale. I wouldn’t quite lump $COHR in with the other photonics players but while they’re already capacity strained, it’s likely going to get worse with Sumitomo, JX affected. This would capacity strain $LITE and others who would lack the inp to produce opticals for $GOOGL TPU Ironwood and others. However this would be offset and we might see a continued rally if -> pricing premium downstream like HBM would offset any product supply shortage. -> Japan manages to remove the export control InP ban in some way (unlikely) -> if $AXTI delivers the substrates at scale to western suppliers (after their $100m funding to double capacity), things will proceed. That being said just wanted to offer another perspective on the drop since I haven’t really seen anyone comment on this from the supply chain angle. I personally removed some optical exposure risk after seeing the supply chain vulnerability this week from export controls and the new monopolistic bottleneck by $AXTI. But I’m not a short seller so I wouldn’t take advantage of it. I’m extremely bullish on the overall segment, but no clue what’s going to happen (whether risk-off selloff continues or rebound) for photonics players.
-
住友受出口管制引发下游产能瓶颈，故减持光子股，COHR例外。
我不做空，也不想发布看空光学公司的帖子，但我短期卖出了其他光子学持仓，因为住友电工(Sumitomo)受到出口管制，因此通过二级效应，$LITE、$AAOI 等下游公司会出现产能问题。 $COHR 算是个奇怪的例外，不确定为什么他们被归入其余公司，因为他们自己生产晶圆。
英文原文
I don’t short sell + didn’t want to publish bear posts about optical companies but I sold my other photonic positions short term because Sumitomo got export controlled so by second order effect $LITE, $AAOI and others would have capacity issues downstream. $COHR is kinda the weird exception not sure why they got lumped in with the rest since they produce wafers themselves.
-
AXTI通过掌控InP供应链实现定价权，涨价对TPU成本影响有限。
^ 澄清一下，$AXTI 掌控定价权。关于数量，指的是掌控磷化铟(InP)供应链的大部分份额。 他们拥有一定量的库存，这显然受中国许可证管制。 重点是，即使产能翻倍或翻三倍，如果你同时掌控供应链，就能掌控定价。你可以用有限的材料将价格提高100%+、500%、2000%，超大规模云服务商(hyperscalers)仍会买单，因为没有替代品。 即便涨幅如此之大，由于此前价格占比极低，对TPU集群物料清单(BOM)的影响也仅增加几个百分点。
英文原文
^ clarification, $AXTI controls the price. Regarding quantity, that was referring to controlling majority of the InP supply chain. They have a certain amount of inventory, which obviously is controlled by China for permits. Point was that can double, triple, capacity, but if you control both the supply chain, you can control pricing. You can have finite material and increase prices 100%+, 500%, 2000%, and hyperscalers would still pay because there's no alternative. Even that much increase would only be a few percent increase on TPU cluster BOM because the price was so negligible before.
-
HBM 瓶颈导致传统建模失效，磷化铟成新瓶颈，$AXTI 因卡位 AI 建设具极高价值。
当出现像 $MU 或 SK 海力士的 HBM（高带宽内存）这样极不常规的瓶颈时，诸如 NTM（下一财年）这样的标准建模就会失效。目前 HBM 的需求极度缺乏弹性，买家不再询问价格，即使在价格飙升数百%后，他们仍在寻求配额分配。在这种情况下，基于 $NVDA 2025 年上半年 EML（边缘机器学习）的现有短缺，以及 $MSFT 等公司仅其项目就需要全球两位数的磷化铟(InP)产量（针对 2026 年下半年至 2027 年），$AXTI 很可能在接下来几个月成为下一个瓶颈。因此，当 $GOOGL 等超大规模云服务商为避免 TPU 项目停滞而迫切寻求配额时，原本数亿 TAM（总可寻址市场）可能迅速膨胀至数十亿甚至上百亿，这使得对瓶颈的建模变得不可能。这本质上是买入“比特分配”、原料控制和瓶颈环节。一家市值 13 亿美元的公司卡住了整个 AI 建设的脖子，在我看来很便宜。我不知道这会走向何方，但我预计会出现极大的价格挤压（我们在 SMM 上已看到迹象），这将增加 $AXTI 的底线利润。许可证波动一直是一个风险因素。但同样，任何延迟都会因为这是单点故障而卡住整个 AI 建设，前提是如果中国继续对日本实施出口管制。
英文原文
Standard modeling like NTM breaks when there's extremely unconventional bottlenecks like HBM with $MU or Sk Hynix. HBM demand is so inelastic rn that buyers aren't asking for price, they're asking for allocation even after hundreds of percent increases. And in this case InP and $AXTI will likely be the next bottleneck in the upcoming months, just based on existing shortages from $NVDA EML H1 2025 and how others like $MSFT require the world's double digit inp output just for their program for H2 2026 into 2027. So it's impossible to model bottlenecks where few hundred million TAM previously might squeeze to few billion or tens of billions when hyperscalers like $GOOGL are desperate for allocation so their TPU program doesn't stall out. This is buying into bit-allocation, feedstock control, and bottlenecks. A $1.3B company bottlenecking the entire AI buildout looks cheap to me. I don't know where this is heading, but I do expect an extremely large price squeeze (we're seeing it now on SMM), which just increases the bottom line of $AXTI. License volatility was always a risk factor. But again, any delays would just bottleneck the entire AI buildout since this is single point of failure if China continues their export controls on Japan.
-
AXTI股价波动属旧账延期，核心应关注衬底瓶颈与InP涨价。
$AXTI 简而言之：人们正对一年前签约收入因延期一个季度而过度反应。此处并未报告前瞻性收入指引。相比于2024年的签约收入，我更关注实时的衬底(Substrate)瓶颈以及磷化铟(InP)价格的上涨。
英文原文
$AXTI TLDR: people are overreacting to contracted revenue from over a year ago being deferred by a quarter. There was no forward revenue guidance reported here. I care more about the real-time substrate bottleneck and increasing InP prices than 2024 contracted revenue.
-
切断关键材料将引发中美全面贸易战，双方受损。
@valuedrift @4drant 英伟达(Nvidia) GPU与美国贸易。如果他们切断关键材料流向美国，将引发一场两败俱伤的全面贸易战。
英文原文
@valuedrift @4drant Nvidia GPUs and American trade. There would be a full blown trade war that hurts both sides if they stop the flow of critical materials to US
-
预计磷化铟短缺带来博弈论配置机会，算法或先猎杀止损。
@dl123abc @william_R2Rclub 嗯，如果我是算法，我会进行止损猎杀(stop loss hunting)——首先触发任何散户的短期恐慌性抛售。 我预计目前会出现严重的磷化铟(InP)短缺，并在未来2-3个季度内出现博弈论类型的资金配置，因此这提供了一个相当不错的机会。
英文原文
@dl123abc @william_R2Rclub Hmm, if I were an algorithm, I’d go stop loss hunting -> trigger any retail short term panic sellers first. Im expecting a major InP shortage around now and game theory type allocations over the next 2-3 quarters, so this presents a pretty good opportunity.
-
指出市场忽视InP瓶颈，基于旧合同抛售存在机会。
@TraderSmooth 市场对磷化铟(InP)当前及未来瓶颈缺乏理解，这提供了一个机会，如果人们是基于2024年或2025年前三季度的递延收入合同而抛售的话。
英文原文
@TraderSmooth The lack of understanding current/future bottlenecks around InP presents an opportunity if people are selling based on deferred revenue from 2024 or q1-3 2025 contracts.
-
材料出口管制致AXTI垄断，下游产能受阻，错杀带来机会
他们确实可以，这一瓶颈向下游传导，损害了 $COHR 到 $LITE 向 $GOOGL 和 $MSFT 的产能爬坡。特别是鉴于住友/JX 面临材料出口管制导致的短缺，$AXTI 将在几个季度后处于垄断地位。人们非理性地对 2024 年合同的前置收益和递延收入确认做出反应，而忽视了当前状况，在我看来这提供了一个绝佳的机会。
英文原文
They can, and this bottleneck flows downstream and hurts $COHR to $LITE capacity ramp into $GOOGL to $MSFT. Especially given $AXTI will be monopoly status in a few quarter once Sumitomo/JX face shortages from material export controls. People illogically reacting to former earnings from 2024 contracts and deferred revenue recognition, while ignoring the situation now presents a great opportunity imo
-
市场误读旧合同数据，忽视当前AI供应链的真实瓶颈。
情况没有变化。这是对2024年及去年第一至第三季度签订的老合同，以及针对一个季度收益递延(Deferred Revenue)的滑稽反应。真正的供应短缺和价格飙升(Squeeze)正在此刻发生。市场通常会提前行动，看着大家还在担心供应短缺全面爆发前的旧季度数据，真是有趣。
英文原文
状況に変化はありません。これは2024年や昨年の第1〜第3四半期に締結された古い契約と、1四半期分の収益繰延べ（デファラル）に対する滑稽な反応です。 主要な供給不足や価格高騰（スクイーズ）が起きているのは、まさに今です。市場は通常、先を見越して動くものですから、供給不足が本格化する前の過去の四半期を気にしている様子は、見ていて面白いですね。
-
指出市场忽视InP短缺挤压，仍纠结于旧合同收入确认。
@Paulyy_K 这个决定由你自行判断。我只是觉得有点好笑，因为目前磷化铟(InP)的价格和短缺情况正面临巨大的挤压。然而，人们却在反应2024年或去年第一季度至第三季度（其中第一季度确认递延）的收入确认和合同。 https://t.co/5ScwrUYBDG
英文原文
@Paulyy_K That decision is up to you. All I’m saying is I’m laughing a bit because what we’re seeing with InP prices and shortages is a huge squeeze. Yet people are reacting to former revenue recognition and contracts made in 2024 or q1-q3 last year (with 1q recognition deferred) https://t.co/5ScwrUYBDG
-
POET 非当前 InP 瓶颈解法，或是 2028 年降 InP 用量的工程方案。
$POET 对于 $AXTI 在 2026-2027 年因磷化铟(InP)瓶颈以及当前专用集成电路(ASIC)/图形处理器(GPU)架构而言，基本不存在。但它是 2028 年以后降低 InP 用量的一种工程变通方案（尽管仍需使用 InP）。例如，富士康 H2 订单良率低、缺乏超大规模数据中心供应商认证，且未纳入 TPU Ironwood 等当前架构设计，目前一切均依赖 InP。但 2028 年推出的未来设计可能会采用它。
英文原文
$POET is basically nonexistent for the $AXTI 1-2 year InP bottleneck for 2026-2027 and current asic/GPU architectures but it’s one of the engineering workarounds to lower InP usage 2028 onward (still requires InP though) Eg. Foxconn h2 orders, low yield, lack of hyperscaler qualification, and not in current architecture designs like TPU ironwood, and everything is currently dependent on InP. But future designs might use it going out in 2028
-
中国新出口管制影响日企前体供应，需二次审视。
@Citrini7 研究信息的“圣杯”，感谢免费分享。此外，由于味之素(Ajinomoto)、日东纺(Nittobo)和三菱瓦斯化学(Mitsubishi Gas Chemical)等公司使用中国原材料前体，其中一些受中国前几天实施的新双重出口管制影响，因此需要进行二次审视。
英文原文
@Citrini7 Holy grail of research info, thanks for posting it free. Also some of these are affected by China’s new dual export controls the other day since Ajinomoto, Nittobo, and Mitsubishi Gas Chemical and others use Chinese raw precursors, so it does require a secondary look
-
出口禁令导致日本基板材料溢价飙升，西方市场定价更高。
这是本地定价。西方市场的定价会有很大不同（会加上溢价），而且现在溢价可能会更加极端，例如由于出口禁令，甚至无法以3000美元的价格为日本基板生产采购材料。此外，在去年12月底禁令实施前，价格就开始飙升了。
英文原文
this is local pricing. Western pricing would be a lot different (premiums applied), and now there'd probably be even more extreme premiums eg. can't even secure materials for $3k in for japanese substrate prod because of the export ban. it also started spiking late Dec last year before the ban.
-
超大规模云服务商竞价推高利基电信产品供应价格
~12月27日只是价格指数级增长的起点，我预计这种上涨趋势将在今年持续。特别是当超大规模云服务商(hyperscalers)试图相互竞价以争夺即时供应时，而该产品的成本本身非常低廉（最初是利基电信产品）。另外，第一张照片是一个基准测试，这张图表更具代表性。
英文原文
~Dec 27th was just just the start of the exponential increase in price, and I expect this to continue increase over the course of this year. Especially as hyperscalers try and outbid each other for immediate supply when the cost itself is so cheap (originally niche telecom product). Also first photo was a benchmark, this graph is more representative.
-
磷化铟供应瓶颈加剧，AXTI成AI基建关键节点。
你正在实时目睹磷化铟(InP)“瓶颈”的演变，而超大规模云服务商正为争夺材料而忙乱。用于光子学的7N级铟(7-N Indium)在SMM上呈抛物线式暴涨，2026年屡创历史新高。这是由$AXTI控制的材料引发的实时“博弈论”场景。
英文原文
You're watching the InP "bottleneck" play out real time while hyperscalers scramble for material. 7-N Indium (required for photonics) is now parabolically exploding on SMM, reaching daily ATHs in 2026. This is a live "Game Theory" scenario for materials controlled by $AXTI. https://t.co/3YDMVfjGTx
-
仅供参考，砷化镓(GaAs)价格在金属导报(SMM)上正在飙升（且在美国市场的溢价更高）。 它是一种关键材料，用于： - 超大规模云服务商的光学设备（TPU集群等） - 光收发器（5G，数据） - 激光雷达(LiDAR)（Robotaxi，无人机，军事） - 光模块（互连集群） - 硅光子(SiPh)激光芯片（英伟达未来的共封装光学(CPO)以及英特尔/博通SiPh引擎使用磷化铟(InP)连续波激光阵列。） $AXTI 的CEO表示他们控制了磷化铟(InP)供应链中40%的份额。并且在日本出口管制后，衬底供应链很可能由他们垄断。 所以一家公司卡住了上述所有环节的脖子。
英文原文
FYI 7n prices are exploding on SMM (and it trades at an even higher premium in US markets) It's a critical material used in: - Hyperscaler optics (TPU pods, etc) - Optical transceivers (5g, data) - LiDAR (robotaxis, drones, military) -Optical Modules (interconnect clusters) - Silicon photonics laser dies (Nvidia’s future co-packaged optics and Intel/Broadcom SiPh engines use InP CW laser arrays.) $AXTI's CEO said they control 40% of the InP majority of the supply chain. And the substrate supply chain would likely be monopolized by them after Japan export controls. So one company bottlenecks everything above.
-
分析$AXTI在InP衬底供应链的垄断瓶颈地位及非对称投资机会。
磷化铟(InP)衬底的可寻址市场(TAM)此前仅为数亿美元，因为它是一种小众的电信大宗商品。 大多数分析师或卖方报告建模错误，因为其增长并非线性。 这是$AXTI在那些渴望获得配额的全球万亿美元级公司之间，对一种关键材料形成的博弈论式供应瓶颈。 你可以看到，一旦$AXTI和InP的供应短缺启动，其远期TAM和利润率将像高带宽内存(HBM)一样呈指数级增长。我之前做过这个评论，以下是总结： 1. 7N铟原料的价格在上海有色网(SMM)开始飙升，我们看到亚洲供应链与西方供应链（$AXTI出口至西方，推高其尚未积压的未来利润率）之间的分化。 12月19日：亚洲金属网报道“铟锭供应商提高美国报价”，标志着西方定价特定分化的开始。 2025年12月24日：随后报道“铟锭供应商提高中国报价”，证实到周末定价压力已在国内和出口渠道固化。 这影响InP衬底生产商的底线，但$AXTI拥有整个成本结构/供应链，因此任何价格上涨仅有利于其利润率。 2. 超大规模云服务商/西方供应链已经为衬底支付溢价。在$NVDA从$LITE、$COHR等公司锁定外延生长激光二极管(EML)产能后，他们已产能紧张。 -> 我们已看到$MSFT和其他超大规模云服务商的部署因产能问题而放缓。 -> 他们将向上游确保材料（此前因是电信大宗商品而极便宜），然后这就像HBM一样变成博弈论式的配额战。但除了$AXTI，你基本上没有其他能大规模生产的公司。 例如，$GOOGL尤其依赖于此，仅$MSFT的Maia项目若全面铺开预计将占用全球两位数的产能，但实际上根本不够。 3. 最关键的是日本InP衬底供应链受到出口管制，当住友/JX方面的材料耗尽时，$AXTI实际上成为最大的InP衬底供应商。 这纯粹是关于在光子AI升级中，你会如何评估单一公司的单点故障价值。我们可以争论语义，但如果它倒下，整个供应链也会随之崩溃。例如，CEO表示在出口管制前（以及他们筹集1亿美元增加产能前），他们占InP供应链的40%。 而这个巨大的整个AI供应链瓶颈仅估值13亿美元，而一些没有实质影响的随机量子公司却价值100多亿美元。 由于没人知道衬底或原料价格会去向何方，很难准确建模那些TAM突然从数亿变为用于每个AI部署的瓶颈和目标价(PT)。 对我来说，这只是买入拥有最纯粹非对称性博弈的瓶颈（现在恰好是垄断）。
英文原文
TAM for InP substrates was few hundred million previously since it was a niche telecom commodity. Most analysts or charters model this wrong since it's not linear. It's a game theory supply bottleneck of a critical material that $AXTI controls between all the world's trillion dollar companies that are desperate for allocation. You can see both forward TAM and margins increase exponentially like HBM once supply shortage kicks in for $AXTI and InP. I made this comment earlier but this is a summary: 1. Price of 7n Indium feedstock has started to spike on SMM and we're seeing bifurcation between asian supply chains + western ( $AXTI exports to western, which inflates their future margins that aren't already backlogged) Dec 19th: Asian Metal reported that "Indium ingot suppliers raise offering prices in United States," marking the beginning of the specific divergence in Western pricing. December 24, 2025: They subsequently reported that "Suppliers of indium ingot raise quotes in China," confirming that the pricing pressure had solidified across both domestic and export channels by the end of the week. This affects the bottom line of InP substrate producers, but $AXTI owns the entire cost structure/supply chain so any price increases just benefit their margins. 2. Hyperscalers/Western supply chains pay premiums for substrates already. They are capacity strained after $NVDA secured EML capacity from $LITE, $COHR, and others. -> We're already seeing slowdown for $MSFT and other hyperscaler deployments likely because of capacity. -> They will go upstream to secure materials (which are extremely cheap, bc earlier they were commodities for telecom), then this just becomes game theory allocation wars like HBM. But instead of sk hynix, mu, samsung, and the others. You basically only have $AXTI left that can do it at scale. EG. $GOOGL especially is heavily dependent on this, and $MSFT alone from maia was est. to take up double digits of the world's capacity if they wanted to ramp up, but there's literally not enough. 3. The biggest thing was that the Japanese InP substrate supply chain got export controlled, effectively making $AXTI the largest InP substrate provider when materials run out on the Sumitomo/JX side. This is just pure what value would you place on single company point of failure on the photonics AI ramp. We can argue semantics but if it goes down the entire supply chain does too. eg. CEO was saying they were 40% of inp supply chain before export controls (and before their $100M fundraise to increase capacity). And this huge bottleneck of the entire AI supply chain is valued at $1.3B, while random quantum companies with no material impact are worth $10B+. It's impossible to accurately model bottlenecks and PTs where the few hundred million TAM is suddenly used for every single AI deployment because nobody knows where the substrate or feedstock pricing goes. For me it's just buying into the bottleneck with the purest asymmetry play there is (which happens to be a monopoly now).
-
AXTI成AI供应链关键瓶颈，预计3-6个月产能告急。
这可能是历史上首次，全球所有万亿美元市值的超大规模云服务商(hyperscalers)都被 $AXTI 这样一家小公司卡住脖子。而在日本对华出口管制禁令后，该公司将在该瓶颈环节占据垄断地位。我不知道市场将如何评估这一瓶颈，我只是单纯指出这一点。你应该自己做研究！不过，我估计在3-6个月内，AI供应链将进入临界状态，当时住友/JX的库存开始耗尽，且超大规模云服务商开始扩产并意识到由于磷化铟(InP)的次级效应，光子组件(photonic components)已无更多产能。仅 $MSFT 一家就预计将在2026年Q4及2027年 Maia 扩产期间占据整个 InP 供应链产出的两位数份额，而他们甚至尚未开始扩产。
英文原文
This is probably the first time in history all the world's trillion dollar hyperscalers are bottlenecked by such a tiny company in $AXTI. And that company would have a monopoly position in that bottleneck post-export China control ban on Japan. I don't know how markets will evaluate this bottleneck, I'm just simply pointing it out. You should do your own research! However, I would estimate that AI supply chains hit a critical state in 3-6 months time when Sumitomo/JX start running out of inventory and when hyperscalers start to ramp up + realize there's no more capacity for photonic components (by second order effect bc of inp). $MSFT alone was projected to take up double digits of the entire InP supply chain output from Maia ramp in Q4 2026, 2027, and they haven't even started to ramp up yet.
-
AXTI垄断InP瓶颈，全球科技巨头受其制约
全球领先的科技巨头因一家小公司而陷入瓶颈，这恐怕是史无前例的。更关键的是，该公司目前在该瓶颈环节占据垄断地位。未来几个月 $AXTI 的走向，以及市场如何评估这一磷化铟(InP)瓶颈，将非常值得关注。
英文原文
世界をリードするテック企業すべてが、これほど小さな一企業によってボトルネック状態に陥るというのは、恐らく史上初めてのことだろう。しかも、その企業は現在、そのボトルネックにおいて独占的な地位を築いている。 今後数ヶ月で $AXTI がどのような結末を迎えるのか、そして市場がこの InP（インジウムリン）のボトルネックをどう評価するのか、非常に興味深い
-
AXTI垄断InP成AI基建单点故障，市场开始关注。
$AXTI 在磷化铟(InP)领域实际上处于垄断地位，且在中国对日本实施出口管制禁令后，成为以下 AI 基础设施建设中唯一的单点故障(Single Point of Failure)环节： - 英伟达 Nvidia ($4.6T) - 微软 Microsoft ($3.59T) - Meta ($1.64T) - 亚马逊 Amazon ($2.57T) - 谷歌 Google ($3.89T) 看来市场开始意识到这一点了。https://t.co/rxGkcHYBtS
英文原文
$AXTI is effectively a monopoly for InP, and the single point of failure for the AI buildout for: - Nvidia ($4.6T) - Microsoft ($3.59T) - Meta ($1.64T) - Amazon ($2.57T) - Google ($3.89T) after China's export control ban on Japan. Looks like markets are starting to find out. https://t.co/rxGkcHYBtS
-
解析中国对日InP禁令细节及美日在关键材料谈判中的不同处境。
这很微妙，但我尽力解释： 1) 这实际上是禁令，不仅仅是出口管制。这涉及任何与军事最终用户或军事用途相关的内容。对于住友电气(Sumitomo Electric)等主要国防承包商来说，这实际上就是禁令。 还有“次级抵制”规则，直接影响住友/JX： 例如，如果另一家公司购买磷化铟(InP)原料，然后试图将原料/成品收发器转让给日本公司，中国可能会禁止相关公司。 这部分非常微妙。如果你想知道为什么美国在这些出口管制下没事，是因为美国允许英伟达(Nvidia)向中国出售H200芯片，作为交换，中国“解锁”了铟和镓等原材料对美国商业科技公司的供应。这是第72号公告，暂停了对美国民用客户“镓、锗和铟”最严格的“原则性禁令”，直到2026年11月27日。但隐含地，微软(MSFT)、Lite等被视为更偏向商业性质（即使它们不是，MSFT确实服务于美国政府），因为这是为了允许中国训练其AI模型而进行的交换。不幸的是，日本没有像美国这样的筹码。 2) 在2026年公告之前，磷化铟(InP)已被列入中国官方的两用物项出口管制清单。金属报告和公司披露确认，InP于2025年2月被置于出口许可证要求之下。 2026年1月关于日本的公告适用于中国两用物项出口管制清单上的所有物品（800-900项）。由于InP已在2025年最终确定为两用物项，因此自动包含在针对日本的新限制中。 $AXTI被允许向受出口管制和光学制造商限制的西方国家发货，但不允许向日本发货。（西方超大规模云服务商本来就是主要目标） 住友将无法获得能够大规模生产成品所需的材料，因为中国控制了全球InP衬底生产所需的大部分关键材料。 所以情况并不完全相同。
英文原文
So this is pretty nuanced but i'll try my best to explain: 1) It's actually a ban. Not just export controls. This refers to anything related to military end-users or for military purposes. For major defense contractors like, Sumitomo Electric, this is effectively a ban. There was also the Secondary Boycott rule, which directly impacts Sumitomo/JX: Eg. if another company buys InP feedstock then tries to transfer the feedstock/finish transceiver to a Japanese company, China could ban the related companies next. This part is really nuanced. So if you're wondering why US is fine under these export controls, US allowed Nvidia to sell H200 chips to China and in exchange, China "unlocked" the supply of raw materials like indium and gallium for U.S. commercial tech companies. This was Announcement No. 72, which suspended the strictest "principle bans" on "gallium, germanium, and indium" for U.S. civilian customers until nov 27, 2026. But implicitly, Microsoft, Lite, and others are treated more as commercial (even they're not, MSFT does serve the US gov), since this was an exchange to allow China train its AI models the critical materials. Unfortunately, Japan has no leverage like the US on this. 2) InP was already added to China's official dual use export control list prior to the 2026 announcement. Metal reports and company disclosures confirmed that InP were placed under export licensing requirements in February 2025. The Jan 2026 announcement regarding Japan applies to all items on China's dual use export control list (800-900 items). Since InP was already finalized as a dual-use item in 2025, it is automatically included in the new restrictions targeting Japan. $AXTI is allowed to ship to Western countries from export controls and optical makers but not to Japan. (Western hyperscalers are main target anyway) Sumitomo would not have the materials able to make the finished products to ship at scale since China controls majority of the world's critical materials required for inp substrate prod. So it's not quite the same position.
-
AXTI若禁运将阻碍AI基建，超大规模云服务商短期仍依赖其供应。
@NEgleston @jaysyoon 如果 $AXTI 无法向美国出口，那么关于人工智能基础设施建设（AI buildout）就有更严重的问题需要担忧。这对美国是一个巨大的威胁，这就是为什么超大规模云服务商（hyperscalers）正在设计绕过该限制的方案（这需要3年时间），在此期间他们仍然处于依赖状态。
英文原文
@NEgleston @jaysyoon If $AXTI is not able to export to US, then there are bigger problems to worry about regarding the AI buildout It's a huge threat to America which is why hyperscalers are engineering around it (which would take 3 years), in the meantime they're dependent https://t.co/Vpq5KQdm13
-
铟价分化及日本出口管制使AXTI成AI光子学关键瓶颈，估值被低估。
这只是我个人的观点。自上次财报电话会议以来，许多情况已发生变化： 1. 7纳米铟(Indium)原料价格在SMM开始飙升，我们看到亚洲供应链与西方供应链（$AXTI向西方出口，推高其尚未积压的未来利润率）之间出现分化。 12月19日：Asian Metal报道“铟锭供应商提高美国报价”，标志着西方定价特定分化的开始。 2025年12月24日：随后报道“铟锭供应商提高中国报价”，证实截至周末，定价压力已在国内和出口渠道固化。 这影响磷化铟(InP)衬底生产商的底线，但$AXTI拥有整个成本结构/供应链，因此任何价格上涨仅有利于其利润率。 2. 超大规模数据中心/西方供应链已为衬底支付溢价。在$NVDA从$LITE、$COHR等公司锁定外延生长(EML)产能后，他们面临产能紧张。 -> 我们已看到$MSFT和其他超大规模数据中心部署放缓，可能源于产能问题。 -> 他们将向上游锁定材料（此前作为电信大宗商品极其便宜），然后这变成类似高带宽内存(HBM)的博弈论分配战。但除了$AXTI，你几乎找不到其他能大规模生产的公司。 例如，$GOOGL尤其依赖于此，仅$MSFT的Maia项目若全面铺开预计将占用全球两位数产能，但实际根本不够。 3. 最大变化是日本InP衬底供应链受到出口管制，当住友/JX方面材料耗尽时，$AXTI实际上成为最大的InP衬底供应商。 这纯粹是关于在光子学AI扩张中，对单一公司故障点价值的评估。我们可以争论语义，但如果它倒下，整个供应链也会崩溃。例如，CEO称在出口管制（及其1亿美元融资扩产）之前，他们占InP供应链的40%。 这个整个AI供应链的巨大瓶颈市值仅为13亿美元，而一些无实质影响的量子公司市值却超100亿美元。 很难对瓶颈进行建模，因为数亿总可寻址市场(TAM)突然用于每个AI部署，且无人知晓衬底或原料价格去向。
英文原文
So this is just my perspective on this. Lot of things have changed since the last earnings call, 1. Price of 7n Indium feedstock has started to spike on SMM and we're seeing bifurcation between asian supply chains + western ( $AXTI exports to western, which inflates their future margins that aren't already backlogged) Dec 19th: Asian Metal reported that "Indium ingot suppliers raise offering prices in United States," marking the beginning of the specific divergence in Western pricing. December 24, 2025: They subsequently reported that "Suppliers of indium ingot raise quotes in China," confirming that the pricing pressure had solidified across both domestic and export channels by the end of the week. This affects the bottom line of InP substrate producers, but $AXTI owns the entire cost structure/supply chain so any price increases just benefit their margins. 2. Hyperscalers/Western supply chains pay premiums for substrates already. They are capacity strained after $NVDA secured EML capacity from $LITE, $COHR, and others. -> We're already seeing slowdown for $MSFT and other hyperscaler deployments likely because of capacity. -> They will go upstream to secure materials (which are extremely cheap, bc earlier they were commodities for telecom), then this just becomes game theory allocation wars like HBM. But instead of sk hynix, mu, samsung, and the others. You basically only have $AXTI left that can do it at scale. EG. $GOOGL especially is heavily dependent on this, and $MSFT alone from maia was est. to take up double digits of the world's capacity if they wanted to ramp up, but there's literally not enough. 3. The biggest thing was that the Japanese InP substrate supply chain got export controlled, effectively making $AXTI the largest InP substrate provider when materials run out on the Sumitomo/JX side. This is just pure what value would you place on single company point of failure on the photonics AI ramp. We can argue semantics but if it goes down the entire supply chain does too. eg. CEO was saying they were 40% of inp supply chain before export controls (and before their $100M fundraise to increase capacity). And this huge bottleneck of the entire AI supply chain is valued at $1.3B, while random quantum companies with no material impact are worth $10B+. It's hard to model bottlenecks where the few hundred million TAM is suddenly used for every single AI deployment because nobody knows where the substrate or feedstock pricing goes.
-
@itsDanielWu 说得在理，我也看到整个行业都在向光子学(Photonics)转型，包括 $LITE 等公司。但在光子学产能爬坡方面，我有关于磷化铟(InP)衬底瓶颈的独立观点，因此我个人对实现规模化持更悲观的态度。
英文原文
@itsDanielWu That's fair, I see the entire industry moving to photonics as well with $LITE and others. But in regards to photonic ramp, I've had my own thesis on InP substrates bottlenecks, so I'm personally more bearish on getting this to scale https://t.co/d6eDk7K204
-
质疑MRVL AEC竞争力，看好CRDO增长，预期光学瓶颈将延长铜缆寿命。
我的意思是，$MRVL 自 2022 年以来难道没有宣布过四次左右像 Alaska A 这样的 AEC（有源铜缆）杀手吗？我把它当作 $AMD 宣称每一代发布最终都要击败 $NVDA 来看待。 $CRDO 的财报和预测一直在以惊人的速度增长，尽管 Marvell 会是严肃的竞争对手（但如果 Marvell 起飞，两者仍有共存空间）。 我认为光学是最大威胁，但我一直预期 $AXTI 的光学组件会出现巨大的光学瓶颈，因此超大规模云厂商很可能被迫延长铜缆的使用寿命。
英文原文
I mean hasn't $MRVL announced an AEC killer with Alaska A like 4+ different times now since 2022? I treat it like $AMD saying they'll finally beat $NVDA with each gen release. $CRDO's earnings + projections just keeps growing at insane rates, though Marvell would be a serious competitor (but room for both if Marvell takes off). I see optical as the biggest threat, but I've been expecting a huge optical bottleneck for opticals components w/ $AXTI, so hyperscalers will likely be forced into copper life extension.