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
-
甲骨文澄清阿比林数据中心进展，回应彭博社引发的抛售。
彭博社关于 OpenAI 和 $ORCL 的报道在周五引发了大规模抛售。 涉及从 $LITE 到 $MU 的半导体供应链。 今天，甲骨文发布声明回应彭博社： “近期关于阿比林（Abilene）站点的媒体报道是虚假且不正确的。 首先，Crusoe 和甲骨文正紧密协作，以破纪录的速度交付全球最大的 AI 数据中心之一。 其次，甲骨文已完成额外 4.5GW 的租赁，以履行我们对 OpenAI 的承诺。” 此前，商务部周四已驳斥了关于 $NVDA 和 $AMD 出口管制的虚假报道。
英文原文
Bloomberg’s piece on OpenAI and $ORCL triggered a massive selloff Friday. Of the semi supply chain from $LITE to $MU. Today, Oracle put out a statement in response to Bloomberg: “Recent media activity about the Abilene site are false and incorrect. First, Crusoe and Oracle are operating in lockstep to deliver one of the world's largest AI Data centers in Abilene at record-breaking pace. Second, Oracle has completed leasing for the additional 4.5GW to deliver on our commitments to OpenAI.” This comes following the Department of Commerce denying the false reports around $NVDA and $AMD export controls on Thursday.
-
光子学是AI新范式，需忍受波动，切勿使用杠杆。
如果有任何变化，我会发帖说明。包括关于 $AAOI 的情况。 不幸的是，目前市场交易主要受宏观因素和恐惧情绪驱动，除非基本面有像 $HIMS 那样极具颠覆性的变化，否则任何基本面消息都会被避险情绪所覆盖。 光子学(Photonics)是人工智能领域的新一轮范式转移，目前只是在等待度过未来一年的波动期。 不幸的是，许多人正经历着这样的回撤，这就是为什么不要使用杠杆(Margin)如此重要。
英文原文
If anything’s changed I would have posted about it. Including with $AAOI. Unfortunately market trades off macro + fear right now and any fundamentals if it’s not extremely game changing like $HIMS will override risk-off sentiment. Photonics is the new paradigm shift with AI, it’s just waiting out volatility over the next year. Unfortunately a lot of people have draw downs like these, which is why it’s important not to use margin.
-
认为地缘政治不会导致科技巨头削减AI资本支出，成本分析优于需求预测。
@nengcunren2527 是的，我同意很多因素在于需求端，但我不认为“七巨头”(Mag7) 或 $GOOGL 会因为伊朗局势紧张而削减资本支出(capex)，从而在人工智能(AI)竞赛中输给了中国。 局势过于错综复杂，在战争升级的短短几天内，还无法对需求进行建模。但成本分析(cost analysis)则更具可操作性。
英文原文
@nengcunren2527 Yep agreed a lot of it is on demand side but I don’t think mag7 or $GOOGL is going to cut capex and lose the AI race to China over Iran tensions. It’s way too multifaceted and can’t model in demand yet from just a few days into war escalation. But cost analysis is more do-able.
-
反驳情绪投资，指出存储涨价驱动指数上涨及利润大增。
仅凭图表就断言“X”的年初至今表现，并引用情绪作为依据，这是最糟糕的投资方式。 股票跟随基本面，如果由创纪录的自由现金流(FCF)支撑，上涨行情才是可持续的。 此次上涨的主要原因主要是SK海力士(SK Hynix)和三星(Samsung)的存储芯片价格飙升，且这两家公司在指数中占据主要权重。 甚至不看DRAM，NAND价格在第一季度(Q1)就上涨了100%以上。 对于第二季度(Q2)，NAND价格在第一季度100%涨幅的基础上再次上涨100%，而分析师此前预测的涨幅仅为5-30%。 这极大地增加了存储芯片制造商的营业利润。
英文原文
Looking at the chart alone for saying "X" YTD and citing sentiment as is the worst way to invest. Stocks follow fundamentals and rallies are sustainable if it's backed by record FCF. The reason for the rally was primarily memory price hikes from Sk Hynix/Samsung and those two making up majority of the index by weight. Not even looking at DRAM, NAND prices were hiked Q1 by 100%+. For Q2, NAND prices again were 100% again, off the Q1 100% hikes while analysts were projecting 5-30% increases. This vastly increases operating income of memory makers.
-
油价/LNG/氦气恐慌被夸大，SK海力士/三星高利润率可吸收成本，逢低买入。
关于韩国综合指数(KOSPI) | $EWY (SK海力士 / 三星) 的情绪， 原油 / 液化天然气(LNG) / 氦气要么： 扰乱供应或压缩利润的说法被夸大了。 对SK海力士、三星的供应链中断和能源成本威胁是耸人听闻的噪音。 原因如下： 1. 原油： 如果油价上涨31%并浮动至120美元/桶的可能情景： 在这种情况下，油价对SK海力士和韩国内存股几乎没有实质性影响。 通过挂钩油价的LNG/JKM价格，韩国股票面临能源成本上升，主要影响那些仅靠5%到10%微薄利润生存的公司。 然而，鉴于内存价格飙升且三星在Q2将NAND价格翻倍，韩国电力公社(KEPCO) 70%的费率上调对三星/SK海力士几乎没有实质性影响。 根据披露的财务数据，SK海力士的年度电费超过1万亿韩元（DIGITIMES报道，约7.5亿美元）。相对于FY2025年97.15万亿韩元的收入，这约占收入的1-2%。 SK海力士在2025年Q4实现了58%的营业利润率。在此背景下，能源成本冲击很小： 如果我们模拟能源成本增加50%： - 对SK海力士季度营业利润(19.17万亿韩元)的影响：~1340亿-~1460亿韩元 (0.76%) - 对三星DS季度营业利润(16.4万亿韩元)的影响：~4070亿韩元 (2.4%) 每次50%的能源成本激增会使SK海力士的利润率减少约0.7%，三星的营业利润率减少2.4%。 分析师预测SK海力士在2026年传统DRAM上的利润率可能达到70%以上。即使能源成本增加100%，也不会对韩国半导体营业利润率构成实质性威胁。 然而，这对营业利润率为5-10%的低利润率公司是实质性的。 输家：传统重工业（钢铁、基础化学品、普通平板玻璃）。 赢家：三星/SK海力士。 主要风险是供应链的二阶效应，如原材料成本增加。这很难建模，但在一个例子中： 一家工业公司迫使原材料（化学品、特种气体）价格上涨30%，这对晶圆厂的影响微乎其微。 材料约占半导体销售成本(COGS)的15-20%，因此从数学上讲，材料成本激增30%只会使SK海力士的营业利润率再减少约2%。 一个打印70%利润率（并提高价格）的寡头可以轻松吸收3-4%的直接（公用事业）和间接（材料）能源逆风。 在大多数情况下，成本可能会通过NAND/DRAM价格上涨转嫁给超大规模云服务商(hyperscalers)。 在油价上涨3倍或5倍的极端情况下。 油价上涨数百倍的主要影响是两个因素： - 全球宏观经济冲击，导致全球通胀（影响从$GOOGL到$COST的每一家公司）。 - 韩元(KRW) USD/KRW汇率崩盘。 持续高油价导致的韩元贬值是一个真实的二阶风险，但历史上韩国内存出口商受益于韩元贬值带来的收入端优势。三星/SK海力士的大部分销售以美元计价，而成本以韩元计价。因此，韩元走弱实际上有利于出口商的利润率，部分抵消了能源成本逆风。 但在油价上涨5倍的极端情况下，那个末日世界里唯一的做多标的只有原油本身、像$LMT / $NOC这样的国防承包商、美国国内能源生产商和美元。 这不太可能发生。 财经媒体和算法可能会恐慌，但如果原油从91美元涨到120美元且KEPCO提高能源成本： 数据显示这对三星 / SK海力士几乎没有影响，主要影响的是那些营业利润率微薄的玩家。 2. LNG： 如果霍尔木兹海峡关闭，韩国大部分LNG进口将不受影响。 媒体一直在引用霍尔木兹海峡+流向中国、印度、韩国和日本的LNG流量。但如果我们查看韩国的贸易数据，这只是其总进口的一小部分。 大部分进口通过非霍尔木兹海峡路线到达，例如澳大利亚 (24.6%)、美国 (12.2%)、马来西亚、印度尼西亚 (~20%) 和俄罗斯/萨哈林 (~4.6%)。其余部分由来自尼日利亚、秘鲁、文莱、巴布亚新几内亚等次要来源填补。 2024年82%的进口是挂钩油价的长期合同，正如我们上面建模的那样，能源成本增加会使运营支出(opex)每增加50%受损1-2%，但鉴于DRAM/NAND价格上涨和营业利润率达到70%以上，这只会造成很小的缺口。 即使如此，成本也会转嫁给超大规模云服务商。 韩国从2022年吸取了教训并多元化了来源，LNG供应中断的影响很小。主要担忧是油价影响LNG价格上涨。 3. 氦气： SK海力士声明：“已确保多样化的供应链和充足的氦气库存”。 “因此，公司几乎不可能受到[氦气]影响。” 现实是，像$TSM到SK海力士这样的大型玩家已经多元化了供应链以应对外部事件。 氦气对半导体供应链至关重要，但媒体叙事是耸人听闻的。尤其是当最大的内存公司发表强硬声明称公司[SK海力士]不可能受到影响时。 _ 但对于三星/SK海力士等韩国股票，围绕石油/LNG/氦气的恐惧看起来与现实脱节： 因氦气和KEPCO费率上调而抛售SK海力士的算法是在进行错误的计算。这从根本上说是一个非供应链问题。 主要威胁是石油和能源成本对全球宏观经济冲击的影响，影响从消费品到通胀的一切。 3月3日“黑色星期二”暴跌使KOSPI下跌7.2%，SK海力士单日下跌11.5%，正是以这些能源安全恐惧为主要催化剂。当然，杠杆强制平仓火上浇油。 然而，基本面与价格走势之间的脱节就是交易机会。如果利润率真的受到威胁，抛售就是合理的。 但，抛售在一天内摧毁的价值比几十年能源成本上涨可能造成的还要多。 数学不支持恐惧的事实正是为什么韩国是买入的理由，因为市场正在因情绪而非结构性扩张的盈利能力（尽管石油/能源成本增加）而抛售。
英文原文
The sentiment around KOSPI | $EWY (SK Hynix / Samsung) on Crude Oil / LNG / Helium either: Disrupting Supply or Compressing Margins are overblown. The supply chain disruption and energy cost threats to SK Hynix, Samsung are sensationalized noise. Here's why: 1. Crude Oil: The a likely scenario if oil prices increase 31% and oil floats to $120/bbl: In this case, the effect on oil has almost no material impact on SK Hynix and South Korean memory equities. There are increased energy costs via oil-pegged LNG/JKM prices on Korean equities, mainly for companies surviving on razor-thin 5% to 10% margins. However, a KEPCO 70% rate hike has little material affect on Samsung/SK Hynix, given memory prices have soared with Samsung doubling NAND prices Q2. From disclosed financial from, their SK Hynix's annual electricity bill exceeds ₩1 trillion per DIGITIMES (~$750M). Which against FY2025 revenue of ₩97.15 trillion represents roughly 1–2% of revenue. SK Hynix posted a 58% operating margin in Q4 2025. Against this backdrop, the energy cost shock is small: If we model a 50% increase in energy costs: - Hit to SK Hynix quarterly OP (₩19.17T): ~₩134 billion-~₩146 billion (0.76%) - Hit to Samsung DS quarterly OP (₩16.4T): ~₩407 billion (2.4%) Every 50% energy cost spike would shave roughly .7% off SK Hynix margins and 2.4% off Samsung operating margins. Analysts project SK Hynix margins could reach 70%+ on conventional DRAM in 2026. Energy costs do not meaningfully threaten Korean semiconductor operating margins, even if they were to increase by 100%. However, this is material to companies with low operating margins of 5-10% The Losers: Traditional heavy manufacturing (steel, basic chemicals, standard flat glass). The Winners: Samsung/SK Hynix. The main risk is second-order effects on supply chains such as increased material costs. This is very hard to model, but in an example where: an industrial company forces 30% price hikes on raw materials (chemicals, specialty gases), it barely dents the fabs. Materials are roughly 15-20% of semiconductor COGS, so mathematically, a 30% spike in material costs only shaves an additional ~2% off SK Hynix's operating margins. A combined 3-4% direct (utilities) and indirect (materials) energy headwind is easily absorbed by an oligopoly printing 70% margins (and increasing prices). In majority of cases, the costs likely get passed down to hyperscalers through NAND/DRAM price hikes. In the very worst case scenario of oil prices increasing 3x or 5x. The main affect on oil increasing hundreds of percent are two factors: - Global macroeconomic shock, causing global inflation (affecting every single company, from $GOOGL to $COST). - KRW (South korean Won) USD/KRW exchange rate blowout. KRW depreciation from sustained high oil is a real second order risk, but historically Korean memory exporters benefit from won weakness on the revenue side. The majority of Samsung/SK Hynix sales are dollar denominated wheras costs are won denominated. So a weaker KRW is actually margin accretive for exporters, which partially offsets the energy cost headwind. But in an extreme case of oil prices hiking 5x, the only longs in that apocalyptic world are crude oil itself, defense contractors like $LMT / $NOC, domestic US energy producers, and the US Dollar. This is unlikely to happen. The financial media and algorithms will likely panic, but , if crude oil goes from $91 to $120 and KEPCO increases energy costs: The data shows there's little affect on Samsung / SK Hynix in specific, and the main impact are on players with razer-thin operating margins. 2. LNG: If the Hormuz closed, the majority of South Korea’s LNG imports would be unaffected. The media has been quoting Hormuz + LNG flows going to China, INdia, SK, and Japan. But if we look at the trade data from South Korea, that’s just a fraction of their total imports. Majority of imports arrive via Hormuz free routes, eg. Australia (24.6%), US (12.2%), Malaysia, Indonesia (~20%), and Russia/Sakhalin (~4.6%). Then the rest filled in with minor sources from Nigeria, Peru, Brunei, PNG, etc. The 82% of 2024 important were long term contracts that were oil-indexed, and as we've modeled above, increasing energy costs would hurt opex by 1-2% per 50% increase, but given DRAM/NAND price hikes and operating margins hitting 70%+, this would make a very little dent. Even if they did, costs would be passed onto hyperscalers. South Korea learned their lesson from 2022 and diversified sources, and there's little impact on LNG supply disruption. The main concern is oil impacting hiking of LNG. 3. Helium: SK Hynix statement: “Long secured diverse supply chains and sufficient inventory" of helium. "Therefore there is almost no chance that the company will be affected [by helium]. The reality is larger players like $TSM to SK Hynix have diversified their supply chains against foreign events. Helium is critical to semiconductor supply chains, but the media narrative is sensational. Especially when the largest memory company puts out an assertive statement that there’s no chance the company [SK Hynix] will be affected. _ But to South Korean equities in Samsung/SK Hynix, fears around Oil/LNG/Helium look disconnected from reality: The algorithms selling off SK Hynix because of helium and KEPCO rate hikes are acting on bad math. It is fundamentally a supply chain non-issue. The main threat is oil and energy costs on global macroeconomic shock affecting everything from consumer goods to inflation. March 3 "Black Tuesday" crash dropped KOSPI dropped 7.2% and SK Hynix fell 11.5% in a single session on exactly these energy security fears as the main catalyst. Of course, forced liquidations from leverage added fuel to the fire. However, the disconnect between fundamentals and price action is the trade. If margins were actually threatened, the selloff would be justified. But, the sell-off destroyed more value in one day than DECADES of hiked energy cost increases could have. The fact that the math doesn't support the fear is precisely why it's Korea is a buy, as markets are selling off on emotion rather than looking at the structural expanding profitability despite increasing oil/energy costs.
-
回应LNG供应担忧，指出长期合同挂钩油价可能推高运营成本。
之前的帖子主要是在回应当时对液化天然气(LNG)供应的担忧。当消息首次发布时，所有人都担心供应中断，导致无法再制造芯片。长期合同很可能与油价挂钩，因此运营支出(opex)可能会因油价上涨而增加。所以仍然需要对石油进行一些建模分析，但这方面因素过于复杂。
英文原文
Post was mainly addressing LNG supply fears back then. When news was first released, everyone was fearing supply shut off and no chips could be made anymore. Long term contracts were likely oil indexed, so there might be opex hikes from oil prices. So still need to do some modeling around oil, which is way too multifaceted.
-
SK海力士澄清氦气供应无忧，驳斥市场恐慌，巨头供应链已多元化。
SK海力士声明： “已长期确保多元化的氦气(He)供应链和充足的库存”。 “因此，公司几乎不可能受到[氦气短缺]的影响。” X上的媒体和散户：“因为氦气，半导体行业完蛋了！” 95%的主流媒体和FinX账号都是纯粹的噪音。 虽然可能会对较小玩家产生影响并带来适度逆风： 但现实是，像$TSM到SK海力士这样的大型玩家已经针对外部事件实现了供应链多元化。他们不会突然停止制造芯片。 尤其是当最大的内存公司发表强硬声明，称公司[SK海力士]几乎不可能受到影响时。 这几乎就像是……世界上最大的公司从2022年吸取了教训？
英文原文
SK Hynix statement: “Long secured diverse supply chains and sufficient inventory" of helium. "Therefore there is almost no chance that the company will be affected [by helium]. Media and Retail on X: “It’s all over for Semis because of Helium!” 95% of mainstream media and FinX accounts are pure noise. While there might be an impact on smaller players and a moderate headwind: The reality is larger players like $TSM to SK Hynix have diversified their supply chains against foreign events. And aren’t going to suddenly stop making chips. Especially when the largest memory company puts out an assertive statement that there’s no chance the company [SK Hynix] will be affected. It’s almost if… the largest companies in the world learned their lesson from 2022?
-
澄清LITE/COHR与中际旭创在光模块供应链中的不同角色。
@LostFundamental 这完全说不通。$LITE 和 $COHR 彼此竞争，但它们与中际旭创(Innolight)完全不同，后者负责设计/组装并从这两家公司采购激光器。它们在光收发器(transceiver)设计上确实有轻微重叠，但其核心业务是激光器。
英文原文
@LostFundamental That makes zero sense. $LITE and $COHR compete with one another but they’re completely different than Innolight that does design/assembly and buys the lasers from the two. They do have slight overlap with transceiver design but their core business are lasers.
-
SK海力士与三星氦气供应安全，市场恐慌错判，属买入机会。
氦气恐慌为 $EWY、KOSPI、台湾和日本制造了引人注目的标题。 这与之前发现韩国巨头安然无恙后的LNG恐慌类似。 然而，底层机制显示，SK海力士和三星等存储厂商（Memory Players）对潜在的氦气中断具有高度免疫力。 SK海力士向路透社发表声明： “已长期确保多元化的供应链和充足的氦气库存。” “因此，公司几乎不可能受到影响。” 三星尚未发表声明，但他们在第四季度财报电话会议中表示： “我们自豪地成为行业内首个开发并部署半导体制造氦气回收系统的公司。” “这使我们能够回收和纯化氦气以供重新部署，每年减少约4.7吨的消耗，并实现约19%的重复使用率。” 三星确认该系统已在生产线上运行，可能保护了其供应链。 当投资者因金永培（Kim Young-bae）关于90%依赖中东进口将导致HBM生产停摆的说法而恐慌时， Volza贸易数据反驳了这一点，但差异可能源于气态与液态氦的分类、转运，或金出于政治效果而夸大其词。 无论如何，这两大韩国巨头似乎已很大程度上确保了其氦气供应链免受中东冲突的影响。 市场抛售的核心盲点在于： 美国不会让其中东冲突阻碍其AI建设。SK海力士和三星正处于这一核心。 美韩技术繁荣协议（US/Korea Technology Prosperity Deal）旨在保护先进技术供应链。 尤其是当像SK海力士这样的公司发表明确声明时： 声明公司不可能受氦气影响。 市场正在为氦气/LNG中断的最坏情况定价，但查看这些声明表明这是一个买入机会。
英文原文
The helium panic makes for a compelling headline for $EWY, KOSPI, Taiwan, and Japan. Similar to the LNG panic before finding out the Korean giants were fine. However, the underlying mechanics show the memory players in SK Hynix and Samsung in specific are highly insulated from potential Helium disruption. SK Hynix put out a statement to Reuters: "Long secured diverse supply chains and sufficient inventory" of helium. "Therefore there is almost no chance that the company will be affected." Samsung have not put out statements yet, but from their Q4 earnings call they stated: "We are proud to be the first in the industry to develop and deploy a helium reuse system for semiconductor manufacturing" This "enables us to recover and purify helium for redeployment, cutting annual consumption by approximately 4.7 tons and achieving a reuse rate of around 19%" Samsung confirmed that this system is already active on production lines and likely protects their supply chain. When investors panic from headlines around structural exposure due to Kim Young-bae's claims about 90% reliance on Middle Eastern imports shutting down HBM production. The Volza trade data contradicts it, but the discrepancy could be gaseous vs. liquid helium categorization, transshipment, or Kim rounding up for political effect. Regardless, it looks like the two large Korean giants have largely secured their Helium supply chain from conflicts in the Middle East. The core point of the selloff markets are missing: United States will not let their AI buildout stall from Middle East Conflicts. SK Hynix and Samsung are in the center of it all. US/Korea Technology Prosperity Deal to protect advanced technology supply chains as well. And especially so, when companies like SK Hynix put out definitive statements: Stating there's no chance the company is affected from helium. Markets are pricing in the worst case scenarios for Helium/LNG disruption but looking at the statements suggest it's a buying opportunity.
-
谷歌CTO暗示AI支出激增，澄清市场误判，看好AI基建加速。
@4intheflames $GOOGL 人工智能首席技术官刚刚明确暗示未来8年将增加支出。 https://t.co/zSf77JAHBQ （该推文引用了 @aleabitoreddit 的推文，引用内容仅供理解语境）： “谷歌数据中心建设规模可能超过1万亿美元” 这对谷歌TPU生态系统1.5万亿美元以上的资本支出意味着什么： 对$LITE、联发科、$AVGO、$TSM、SK海力士、三星等均为广泛利好。 在接受福布斯关于人工智能资本支出的采访时，$GOOGL首席技术官表示： “今年（资本支出）在1750亿至1850亿美元之间，假设不会下降，人们可以想象这在10年内可能延伸到某个巨大的数字。” 福布斯评论道：“谷歌的数据中心野心与OpenAI有着巨大差异：谷歌是一台赚钱机器。” 昨天，市场抛售了OpenAI和$ORCL相关的半导体股票，导致一些“星际之门”相关项目暂停。 市场将OpenAI的成长阵痛与由最富有的超大规模云服务商资助的更广泛的人工智能基础设施建设混为一谈。 从$GOOGL和印钞机般的Mag7角度来看： 人工智能基础设施建设在未来十年看来只会加速。
英文原文
@4intheflames $GOOGL CTO of AI literally just signaled increase spend over the next 8 years. https://t.co/zSf77JAHBQ
-
NVDA锁定EML产能，光子学上游材料受制于中国，COWOS/HBM/光子学成瓶颈。
$NVDA 去年锁定了大部分光模块(EML)产能，导致其他超大规模云服务商面临巨大瓶颈。然而，光子学整体是一个材料问题，最上游的材料由中国控制。这并非英伟达能单方面决定的。不过，他们通常是最早布局的，这给 $META、$MSFT、$AMZN、$GOOGL 留下了更多问题，因此本文指出共封装光学(COWOS)、高带宽内存(HBM)和光子学可能是瓶颈。
英文原文
$NVDA secured most of the EML capacity last year and caused a huge bottleneck for other hyperscalers. However, photonics as a whole is a materials problem, the most upstream materials controlled by China. Not exactly something Nvidia can decide. However, they're usually the first to things and that leaves more problems to $META, $MSFT, $AMZN, $GOOGL afterward, hence the point of the post signaling cowos, hbm, and photonics are probably the bottlenecks.
-
指出MRVL在光子学领域的布局，但强调其收入主要依赖ASIC。
@cesarsroy 是的，$MRVL 在光子学领域（如与 Celestial 合作的共封装光学(CPO)）有很多布局。上面应该提到他们的，但我总是把他们和专用集成电路(ASIC)联系在一起，因为那是他们大部分收入的来源，比如 2027 年的 $MSFT Maia。
英文原文
@cesarsroy Yes $MRVL does a lot of stuff in photonics like cpo with celestial. should have included them above, but I always just associate them with asics since that's where majority of their revenue comes from like $MSFT maia in 2027.
-
光模块或随超大规模云厂商锁定产能而结束低价，面临涨价预期。
据我所知，自超大规模云服务商认为800G/1.6T光模块已是大宗商品(commodities)并提前获得批量折扣(volume discounts)以来，价格并未大幅上涨。但这类似于内存(memory)转变为结构性需求，加上超大规模云服务商为保障产能而带来的极端需求。如果看到它们开始涨价，我不会感到惊讶。
英文原文
There haven't been much price hikes to my knowledge since hyperscalers were thinking 800g/1.6T are commodities + getting volume discounts earlier. But similar to memory transforming to something structural + extreme demand from hyperscalers securing capacity. Would not be surprised if we see them start hiking prices.
-
澄清EWY为存储股，指出NAND涨价超预期但属周期中段，光子学尚早。
不，$EWY 主要包含三星和 SK 海力士，这两家主要是存储芯片厂商。 三星昨日宣布第二季度 NAND 闪存价格上调 100% 以上，而分析师此前普遍预测仅比第一季度增长 5-30%。 NAND/DRAM 是一个独立的瓶颈环节，但我们正处于周期中段（尽管需求可能是结构性的，但季度价格暴涨 100% 的可能性不大）。 光子学(Photonics)仍处于早期阶段。
英文原文
No, $EWY is Samsung/SK Hynix which are primarily memory. Q2 NAND prices were hiked 100%+ by Samsung yesterday while analyst expectations were all projecting 5-30% increases from Q1. NAND/DRAM is a separate bottleneck but we're midcycle (but demand is likely structural, 100%/quarter price hikes, probably not). Photonics is early.
-
超大规模云厂商资本支出未放缓，光学组件产能已满，预示AI光子学瓶颈将至，建议持有。
情绪（如战争/石油/LNG）决定短期价格，基本面决定长期价格。 我看不到那些关键的超大规模云服务商（$GOOGL, $MSFT, $AMZN 等）的资本支出有任何放缓迹象。 而且，在光学组件全面扩产之前，从 $AAOI 到 $LITE 的所有产能已在涨价前满负荷运转。 当新的 AI 集群架构采用光子学进行扩展时，这就是即将出现的瓶颈的定义。 如果这些公司的基本面没有发生任何破坏性变化，可能最好持有以度过短期波动。
英文原文
Sentiment (eg. War/Oil/LNG) dictates short term prices, fundamentals dictate prices in the long run. I see zero slowdown from the hyperscaler capex that matters ( $GOOGL, $MSFT, $AMZN, etc). And before the ramp for optical all their capacity from $AAOI to $LITE are already maxed out before price hikes. This is the definition of an upcoming bottleneck when new AI cluster architectures use photonics to scale. If there's been nothing breaking the fundamentals of any of these companies, probably better to hold through short term volatility.
-
光子学成AI新瓶颈，激光与InP紧缺，CPO变革预计2028年。
光子学是下一个主要瓶颈。 $NVDA 提前发出了每一个瓶颈的信号：从 HBM（与三星/SK海力士合作）到 CoWoS，再到如今对 $LITE 和 $COHR 的投资： 激光晶圆厂、共封装光学(CPO) 和磷化铟(InP)。 针对每个瓶颈中最具非对称性的多头标的： 1. InP 衬底：$AXTI，住友，JX 2. InP 上游原料+加工：$AXTI 3. 激光器：$AAOI（自研），$AVGO，$COHR，$LITE 4. CPO：$TSEM，Soitec。 $AAOI 财报电话会确认了激光器瓶颈，三家超大规模云厂商希望买断其能生产的所有光收发器。 $AXTI 的积压订单确认了 InP 衬底瓶颈。（玩家图片来源：IndexBox） CPO 瓶颈被广泛预期将在 2027-2028 年末发生。 $AVGO 关于“CPO”的评论带来了短期波动。但这不同于正在发生的激光->收发器和 InP 瓶颈。 关于时间框架： $AAOI，$LITE，$COHR 和激光收发器瓶颈正在实时发生（并预计像内存一样在 2028 年前恶化）。 $AXTI，住友和 InP 衬底瓶颈正在实时发生（并预计只要 AI 在未来多年使用光子学，情况就会恶化）。 由 $NVDA 引领的向 CPO 的大型架构转变可能发生在 2028 年。 这些对于 AI 的下一个范式转变来说似乎是不可避免的。
英文原文
Photonics is the next major bottleneck. $NVDA has signaled each one ahead of time from: HBM (with Samsung/Sk Hynix) to CoWoS and now with the $LITE and $COHR investment: Laser Fab, CPO, and InP. For the most asymmetrical longs in each bottleneck: 1. InP Substrates: $AXTI, Sumitomo, JX 2. InP Upstream Feedstock + Processing: $AXTI 3. Lasers: $AAOI (internal), $AVGO, $COHR, $LITE 4. CPO: $TSEM, Soitec. The laser bottleneck was confirmed from the $AAOI earnings call when three different hyperscaler wanted to buy out any optical transceiver they can produce. The InP substrate bottleneck was confirmed with the backlog from $AXTI. (Image source of players: IndexBox) And the CPO bottleneck is widely expected to happen later in late 2027-2028. There's short term volatility from $AVGO comments around "CPO" in specific. But that's different than the laser -> transceiver and InP bottlenecks happening now. For timeframes: $AAOI, $LITE, $COHR and the laser transceiver bottleneck is happening real time (and is expected to get worse like memory into 2028). $AXTI, Sumitomo and the InP substrate bottleneck is happening real time (and is expected to get worse as long as AI uses photonics for the many years to come). And the larger architectural shift to CPO led by $NVDA will likely happen in 2028. These feel inevitable for the next paradigm shift in AI.
-
谷歌AI基建支出庞大且持续加速，利好相关半导体供应链。
“谷歌数据中心建设规模可能突破1万亿美元” 这对谷歌TPU生态系统1.5万亿美元以上的资本支出(capex)意味着什么： 对$LITE、联发科(Mediatek)、$AVGO、$TSM、SK海力士、三星等普遍利好。 在接受福布斯关于AI资本支出的采访时，$GOOGL首席技术官(CTO)表示： “今年在1750亿至1850亿美元之间，假设不会下降，人们可以想象这在10年内可能延伸到某个巨大的数字” 福布斯评论道：“谷歌的数据中心野心与OpenAI有着巨大差异：谷歌是一台赚钱机器” 昨日，市场抛售了OpenAI和$ORCL相关的半导体股，暂停了一些Stargate相关项目。 市场将OpenAI的成长阵痛与由最富有的超大规模云服务商(hyperscalers)资助的更广泛的AI基础设施建设混为一谈。 从$GOOGL和印钞机般的Mag7角度来看： AI基础设施建设在未来十年看来只会加速。
英文原文
“Google’s Data Center Buildout Could Top 1 Trillion” The implications for $1.5T+ in capex spend for the Google TPU ecosystem: From $LITE, Mediatek, $AVGO, $TSM, SK Hynix, Samsung and others are widely positive. In an interview with Forbes on AI capex spend, $GOOGL CTO stated: “We’re at $175 to $185 billion this year, one could imagine, assuming it's not going to go down, that this could extend to some big number over 10 years” Forbes comments: “There’s a big difference between Google’s data center ambitions and OpenAI’s: Google is a money-making machine” Yesterday, markets sold off semi names from OpenAI + $ORCL putting a pause on some Stargate related projects. The market conflated OpenAI's growing pains with the broader AI infrastructure buildout funded by the richest hyperscalers. From $GOOGL and a money-printing Mag7 perspective: The AI buildout only looks to accelerate over the next decade.
-
是的，如果你向光模块的光学收发器更上游追溯，例如 $AXTI 是光子学领域最终的材料瓶颈。他们通常供应给外延片(epiwafer)公司如 $IQE，后者再供应给 $LITE -> $LITE Cloudlight -> $FN -> $GOOGL。沿途还有像 $TSEM 这样的公司，以及向 $IQE/Landmark 出售反应器的随机欧洲厂商。
英文原文
Yeah if you go more upstream for optical transceivers for example, $AXTI is the end of the end materials bottleneck for photonics. They usually they supply to epiwafer companies like $IQE that supply to $LITE -> $LITE cloudlite -> $FN -> $GOOGL. And companies like $TSEM along the way + random European reactor sellers to $IQE/Landmark.
-
最大回报源于亚洲供应链瓶颈及美系利基股，而非英伟达。
@pepemoonboy 经常看到这张图，所以花了很多时间做了一张新的！ 但这确实是事实，最大的回报大多来自亚洲的随机瓶颈环节，如日东纺(Nittobo)、宏碁(Macronix)和南亚玻璃(Nanya Glass)，以及美国玩家如 $SNDK 和 $LITE，而不是 $NVDA。
英文原文
@pepemoonboy Kept seeing this image pop up a lot so spent a lot of time to make a new one! But it's genuinely true though, most of largest returns came from random bottlenecks in Asia like Nittobo, Macronix, and Nanya Glass to US players like $SNDK and $LITE instead of $NVDA.
-
真正超额收益来自控制铲子供应商所需的高价材料瓶颈。
当市场说“当所有人都去挖金矿时，卖出铲子”时，他们错了。为什么？真正的超额收益来自：对铲子供应商进行瓶颈控制。从$AXTI和$LITE的光子学衬底/EML激光器，到$SNDK向SK海力士供应NAND/DRAM，这些都是瓶颈环节。它们最终比那些聚集在$NVDA等铲子供应商周围的投资者更赚钱。这句谚语应该是：“当所有人都去挖金矿时，通过铲子供应商需要的高价材料来获利”。
英文原文
Markets have got it wrong when they say: "When Everyone Digs for Gold, Sell Shovels" You know why? The real alpha comes from: Bottlenecking the Shovel Sellers. Substrates/EML lasers in photonics from $AXTI and $LITE. to $SNDK to Sk Hynix for NAND/DRAM are those bottlenecks. And they end up individually more profitable, than the investors who crowd around shovel sellers like $NVDA. The saying should go: "When Everyone Digs for Gold, Profit off the Materials at jacked up prices that the Shovel Sellers need to Sell the Shovels".
-
对比光通信私有公司高估值与上市公司基本面，指出私有公司若上市恐崩盘。
在光通信(CPO)领域，从 $LITE 到 $AAOI 的内幕人士抛售行为，与公司创纪录的收入/利润形成实质性差异。与此同时，私有公司估值翻倍（例如从 $910亿涨至 $1590亿面向散户），而 $HOOD 到 $COIN 等上市公司却在暴跌。如果这些私有公司上市，其虚高的估值很可能导致股价崩盘，而上市公司基本面依然稳健。
英文原文
Insider selling on photonic names from $LITE to $AAOI while companies generate record revenue/profit is materially different. Private companies hike valuations by 2x (eg. $91B -> $159B to retail) while public companies from $HOOD to $COIN are crashing. If these companies traded publicly, most would all likely tank from their inflated valuations, while public companies are fundamentally sound.
-
驳斥NVDA/AMD出口管制误报，指出媒体煽动性报道影响算法。
我相当确定两者之间存在某种中间地带，甲骨文(Oracle)可能会推迟部分计划，而OpenAI/甲骨文通过CNBC发布消息表示建设仍在继续。 彭博社(Bloomberg)经常发布极其耸人听闻的文章，算法通常会随之跟进。 不过，昨天关于$NVDA和$AMD的虚假出口管制报道尤其恶劣。
英文原文
Pretty sure there’s some middle ground between the two with Oracle maybe delaying some parts of it while OpenAI/Oracle via CNBC putting out that the buildout is continuing. Bloomberg runs extremely sensational articles and algorithms usually follow that. Yesterday’s false export control reporting on $NVDA and $AMD was especially egregious though.
-
澄清Oracle数据中心项目正常，视AI板块错杀为买入机会。
针对彭博社报道的最新回应：“Oracle与OpenAI的数据中心项目现有计划仍在按部就班进行。” 昨日因虚假的出口管制新闻，从$NVDA到$LITE的半导体板块遭遇暴跌。现在，关于$ORCL的消息是否会导致又一个类似的交易日？
英文原文
In a new response to Bloomberg article: “Existing plans for Oracle's data center project with OpenAi remain on track” The first Bloomberg report caused a crash in Semis from $NVDA to $LITE yesterday from false export control news. And now we have potentially another one day regarding involving $ORCL?
-
三星NAND涨价远超预期，利好存储板块。
据《证券日报》报道，“三星电子(005930)据报道将在第二季度将其NAND闪存(NAND Flash)供应价格翻倍”。 三星（以及可能跟随的$SNDK和SK海力士）的NAND涨价幅度几乎是某些分析师预期的20倍，是其他分析师预期的2-4倍。 根据法国巴黎银行(BNP Paribas)分析师3月的笔记，他们预测NAND在Q2环比增长5%。 据ITJungle：“行业分析师预计Q2较Q1上涨20%至50%，Q3上涨5%至15%，Q4上涨5%至10%。” 麦格理(Macquarie)对SK海力士约1905亿美元营业利润的激进预测开始看起来更有道理了。 这对存储公司的影响是惊人的。
英文原文
"Samsung Electronics (005930) will reportedly double its NAND Flash supply price in the second quarter" - Sedaily The price hike of NAND from Samsung (and likely $SNDK + SK Hynix following) is almost 20 times some analyst expectations and 2-4X others. Per BNP Paribas Analyst note in March, they projected for NAND: a 5% Q/Q increase in CQ2 Per ITJungle: "The industry analysts have Q2 up over Q1 in a range of 20 percent to 50 percent, and Q3 at 5 percent to 15 percent, with Q4 up 5 percent to 10 percent" Macquarie's wild SK Hynix ~$190.5 billion operating income projections is starting to look a little more reasonable. The implications for memory companies is crazy.
-
美商务部辟谣AI扩散规则回归，称算法抛售半导体报道不实。
美国商务部发布报告回应路透社和彭博社：“今天有报道称我们将恢复人工智能扩散规则。我们不会。该规则负担沉重、过度越权且后果灾难性。” 此前关于 $NVDA 和 $AMD 的报道，称算法正在抛售半导体股票，从头到尾都是虚假的。
英文原文
US Department of Commerce put out a report in response to Reuters and Bloomberg: “Today there was reporting that we were returning to the AI diffusion rule. We will not. It was burdensome, overreaching, and disastrous.” The reporting around $NVDA and $AMD, that algorithms have been selling off semis on was false all along.
-
驳斥彭博社误报出口管制草案引发半导体供应链恐慌性抛售。
彭博社刚刚引发了半导体市场的崩盘。凭记忆，受影响的标的从 $EWY 和 $SNDK 到 $NVDA 和 $TSM 都有。起因是假新闻。这简直是路透社关键矿产误导性报道的2.0版本。肯定得有人为此负责？
英文原文
Bloomberg just caused a semiconductor market crash. From memory names in $EWY and $SNDK to $NVDA and $TSM. Off fake news. This is Reuters critical minerals misleading reporting v2. Certainly someone needs to be held accountable? https://t.co/YaBNpl8nFB
-
澄清AVGO推迟CPO利好传统光模块，市场误读导致板块错杀。
整个市场都误解了 $AVGO 首席执行官关于共封装光学(CPO)的言论。 这导致今天光子学板块从 $LITE 到 $COHR 在早盘出现抛售。 市场往往根据头条新闻进行交易，却忽略了所有细微差别。 CPO 是光子学技术的一种演进。 这种延迟对某些光子学参与者如 $AAOI 来说极其利好。 虽然对 Soitec 等 CPO 公司而言总体是负面的。 市场和人们将 CPO 与整个光子学板块混为一谈。 Tan 基本上只是给传统光收发器制造商的产品生命周期带来了巨大的延长。 原话如下： “我想说的是，你不需要去追逐那些被称为 CPO 的闪亮新事物，即使我们是 CPO 的领导者。CPO 会在适当的时候到来，不是今年，也许也不是明年，但会在它该来的时候到来。” 事实上，$GOOGL 拒绝 CPO 且 $AVGO 将 CPO 时间线预测推迟数年： 这可能是对许多光子学公司（如开盘下跌 8% 后翻红的 $AAOI（光收发器））来说最棒的事情。 机架内部，超大规模数据中心仍可使用铜缆。但对于更长距离，使用的是可插拔光收发器（现在其生命周期延长了数年）。 个股往往被一起抛售，但最大的赚钱机会在于理解大家忽略的细微差别。
英文原文
The entire market misunderstood $AVGO CEO’s comments around CPO. This caused a selloff today on the photonics sector from $LITE to $COHR in the morning. Markets trade off headline news but miss all the nuances. CPO is an evolution to photonics. This delay is extremely bullish to certain photonics players like $AAOI. While net negative to CPO companies like Soitec. Markets and people conflated CPO with the photonics sector as a whole. Tan basically just gave traditional transceiver makers a massive extension on their product lifecycle. The exact quote: "What all I'm trying to say is you don't need to go run into some bright shiny objects called CPO, even as we are the lead in CPOs. CPOs will come in its time, not this year, maybe not next year, but in its time." In fact, $GOOGL rejecting CPO and $AVGO delaying CPO timeline projections by years: Is probably the most greatest thing that can happen to many photonics companies like $AAOI (optical transceivers) that dropped 8% on open then recovered to green. Inside the rack, hyperscalers can still use copper. But for longer distances, pluggable optical transceivers are used (and now get a life extension by years). Names get sold off all together, but where the most money to be made is understanding the nuances everyone misses.
-
澄清博通CPO评论，指出市场误将CPO利空等同于光子学板块利空。
是的，光子学板块今天被抛售，我的持仓也有轻微回撤，但我觉得人们误解了 $AVGO 关于共封装光学(CPO) 评论的细微差别。 这对许多光模块(optical transceiver)个股其实是非常看涨的。比如 $AAOI 先跌了 9% 但最终翻红。 主要是像 Soitec 这样具有 CPO 敞口的个股会是净利空。 但大多数人甚至算法都将 CPO 与光子学混为一谈，导致整个板块被抛售。
英文原文
Yeah photonics got dumped today, had a slight drawdown too, but feel like people missed the nuance for $AVGO CPO comments. It’s actually extremely bullish for a lot of the optical transceiver names. Stuff like $AAOI sold off 9% then ended up green. Mainly cpo exposure names like soitec would be net negatives Majortity of people and likely algos conflate cpo with photonics though so whole sector got sold off
-
机构对小众光子上游认知滞后，X平台热度具误导性。
对于 $AAOI 来说并不意外。许多机构可能直到被大型 X 账号提及，才听说过像 $AXTI 这样更小众的光子学上游玩家。FinX 社区对这些标的介入极早，但由于 X 平台本身存在泡沫效应，感觉似乎已经“拥挤”。但如果你问这里以外的人这些名字是什么，他们完全一无所知。
英文原文
Not surprising with $AAOI. A lot of institutions probably never heard of a lot of the more niche photonic upstream players like $AXTI until they get mentioned by larger X accounts The FinX community is extremely early to a lot of these names but it feels “crowded” because X is a bubble. But if you ask anyone outside of here what some of these names are, they’ll have 0 clue.
-
光子学板块普跌，小市值公司波动性显著高于巨头。
@CryptoVandelay 一个很好的教训是，如果像 $IQE 这样的股票单日能上涨 25%，它也可能在单个交易日内出现 20% 的回撤。整个光子学(Photonics)板块都在下跌，$COHR 下跌了 8%。鉴于它是一家市值 450 亿美元的公司，我认为较小的公司波动性更大。
英文原文
@CryptoVandelay A good lesson is if a stock like $IQE moves up 25% a day, it can also have 20% drawdowns in a single session too. Whole photonics sector is down, $COHR is down 8%. Given it’s a $45B company, I’d say smaller ones are more volatile.