供应链分析

产业链结构、上下游与瓶颈环节研究 · 共 1122 条 · 第 7 / 23 页

  1. Amazon 的 warrant 协议会带来放量

    @PhotonCap 基本上每次 $AMZN 有像 $ALAB 2024、$AAOI 2026 这种 warrant 协议时,量产订单都会起飞。

    英文原文

    @PhotonCap Basically every time $AMZN has these warrant agreements like $ALAB 2024, $AAOI 2026, volume orders go brrr

  2. SIVE 的终端用户是谁

    以防你还在想 $SIVE 的终端用户到底是谁: Sivers 激光 -> $AMZN(Marvell 的 8-K SEC 文件里有) Amazon 和 Celestial 之间有 “photonic fabric” 的采购协议,而这条链通过两跳关系映射到 Sivers 激光。 协议里 Marvell 的股份价值按 87 美元/股算,所以 Amazon 有很强的动力尽可能多买(Marvell 现在股价已经 139 美元以上了)。 这些量产订单对 $SIVE 最有帮助,因为他们体量小。 我只是想指出这一条关系。 但在 7.2 亿美元市值下…… 市场显然漏掉了 Sivers 作为未来超大规模云光学超级周期里关键激光供应商的角色。

    英文原文

    Just in case you were wondering who the end users of $SIVE likely were: Sivers lasers -> $AMZN (Marvell's 8-K SEC filings). Amazon has purchase agreements for "photonic fabric" from Celestial, which maps to Sivers lasers through two-hop connections. Marvell's share value under the agreement was $87/share, so Amazon is very incentivized to buy as much as they can (Marvell is trading at $139+). These volume orders help $SIVE the most out of this supply chain, given their size. This is just one-relation to Amazon, I wanted to point out. But at $720m MC... Markets missed Sivers as the critical laser supplier for the upcoming hyperscaler optical supercycle based on new architectural changes.

  3. 台湾的封装与电网瓶颈

    @aiizisw77940 我发了很多新的名字,比如台湾的 Shunsin,用来做光学组装 / 封装。 然后 $HSP.A 作为数据中心变压器瓶颈,未来一年大概率也会兑现(因为这类瓶颈是多年的)。

    英文原文

    @aiizisw77940 I posted a lot of new ones like Shunsin for optical assembly/packaging in Taiwan. Then $HSP.A for DC transformers bottleneck will likely play out over the next year (since they have a multi-year bottleneck).

  4. 台湾/韩国/Foundry 受益面概览

    简单给你个 TLDR,帮你省时间: 1. $ASML、$TSM 财报 = 前景不错。半导体 + capex 起飞。 2. Opus 4.7 + Anthropic 起飞。软件股就很惨。 3. 三星起飞,部分原因是 $TSLA 的 AI 芯片。 4. $UMC = 代工涨价。foundry 起飞。 5. 训练在中国起飞。H100 租赁价格上升,neocloud 也开心。 6. 氦气短缺 = 影响不大……我之前已经说过很多次了,但我不确定 $TSM 还要再说多少遍。 7. MLCC、电感价格 = 涨价。受益者我后面再展开。 8. “台湾的 OSAT 扩张可能会收紧全球测试产能并推高成本” 我最近做多了像 Shunsin(6451)这样的台湾 OSAT,有原因的。需求会继续跑赢供给,即使扩产也是一样。(cowos、sip、光学)

    英文原文

    Just some TLDRs to save you time: 1. $ASML, $TSM earnings = Good Outlook. Semis + capex go brrr. 2. Opus 4.7 + Anthropic go brrr. Software = sad. 3. Samsung go brrr because of partly bc of $TSLA AI Chips. 4. $UMC = price hike for foundry. foundries go brrr. 5. Training = brrr in China. H100 rental increase go up. Neoclouds happy. 6. Helium supply shortage = not significant... I've already said this before, but I'm not sure how many times $TSM needs to say this. 7. MLCC, inductor prices = price hike. Will cover beneficiaries later. 8. "Taiwan's OSAT expansion could tighten global test capacity and raise costs" I went long on Taiwan OSATs recently like Shunsin (6451) for a reason. Demand will just outstrip supply, even after expansion. (cowos, sip, optical).

  5. Anthropic终于强制KYC验证,建议美国AI前沿实验室统一推行以阻止蒸馏泄露。

    我去,终于Anthropic听进去了并强制推行KYC验证。早该如此了,别再让DeepSeek、月之暗面(Moonshot)和MiniMax免费蒸馏Opus和其他最新模型了。(引用推文内容:今天Anthropic点名了DeepSeek、Moonshot和Minimax的蒸馏行为。OpenAI也面临同样的问题,被中国实验室薅羊毛。最简单的解决方案没人做:KYC验证。每个金融机构从$HOOD到$IBKR都这样做,以追踪资金流向(如Persona自拍/身份证验证)。同样的方法应适用于token流向。美国前沿实验室应尽快对最新模型实施KYC。我确信他们关心无摩擦增长,但这关乎国家安全,个人和企业只需完成一次这个流程即可。这是目前为止最容易实现的方案,一天就能完成。如果不这样做,中国实验室将继续寻找绕过方法。)

    英文原文

    Holy **** finally Anthropic listened and forced KYC. About time, stop letting DeepSeek, Moonshoot, and MiniMax have a free pass to distill Opus and other latest models. https://t.co/Ba8XBTIsnb

  6. 批评别人误判为拉盘砸盘,不懂超大规模云供应链。

    @Broasted_King @KobeKapital 哈哈,那家伙居然还觉得这是“pump and dump”或者有泄密。 他们想怎么甩锅都行,但欧洲那帮人莫名其妙地这么酸,真离谱。 尤其是他们根本不懂 hyperscaler 的供应链是怎么运作的。

    英文原文

    @Broasted_King @KobeKapital Lmfao that guy is regarded thinking it’s a “pump and dump” or there’s a leak. They can point fingers all they want, somehow people in Europe are extremely salty for no reason. Especially when they don’t understand how hyperscaler supply chains work.

  7. 感叹本该买些,现在后悔它涨起来,因为它做 AVGO 激光模块。

    @StrateGeee 啊,我本来应该买一点的……现在看着它涨上去有点后悔,因为他们给 $AVGO 做激光模块之类的东西。

    英文原文

    @StrateGeee Ah I should have bought some… I have regrets seeing it go up now since they do the laser modules and stuff for $AVGO

  8. 看好 PCL Technologies 借助 AVGO 光子生态翻倍。

    我其实没有持仓,但前面已经给订阅者做过一些简短 DD。 我预计 PCL Technologies(4977)也会翻倍。 因为他们在 $AVGO 的 photonics 生态里负责组装、光学、测试等全部环节。 我记得我在它大概 4.5 亿美元市值的时候就提过这点? https://t.co/vmgilSKANV

    英文原文

    I actually don’t have positions but gave some slight DD away for subscribers earlier. I expect PCL Technologies (4977) to double as well. Since they’re doing everything in $AVGO photonics ecosystem for assembly, optical, testing. Think I mentioned this around ~$450M MC? https://t.co/vmgilSKANV

  9. 详细拆解 Shunsin 的 Foxconn 光学封装/组装定位与利润率提升。

    Shunsin(6451)。鸿海的光学封装/组装业务。 此前作为 OSAT 的毛利率只有 15-16%,但 CPO/1.6T 可能做到 30-35%,这会把整体利润率拉上去。 因为他们在 $NVDA 的 CPO 供应链里做更难的 FAU 光学对准。 和 $FN 或 $ASX 这些市值动辄几十亿美元的公司比呢? 这家公司现在才 14 亿美元市值,我个人推演的 2028 年 forward P/E 可能会从很低的水平压缩到个位数,因为鸿海会把订单转给它。 鸿海体量巨大,而且还能帮这家公司去风险。 我对它很有信心,觉得很快就会翻倍。 只是我认为几乎没人知道这家公司。

    英文原文

    Shunsin (6451). Foxconn’s optical packaging/assembly arm. Previously 15-16% gross margins as OSAT but CPO/1.6T could be 30-35%, which brings blended up. Since they’re doing harder FAU optical alignment in $NVDA CPO supply chains. Compared to $FN or $ASX that trades in the tens of billions? This trades at… $1.4B MC with, fwd 2028 p/e likely compresses to single digits from personal projections, as Foxconn passes orders to them. Foxconn is massive. And de-risks this company. Pretty high conviction this ends up doubling soon. Just nobody knows about this company imo.

  10. 认为富士康光学供应链比无营收试制公司更有价值。

    @StrateGeee 鸿海这边是活生生的……$NVDA 的光学供应链,负责封装、测试和组装…… 居然比一个没有收入的 $LWLG 和 $TSEM 之间的开发试验还便宜。

    英文原文

    @StrateGeee Foxconn’s live.. $NVDA optical supply chain for packaging, test, and assembly.. Is valued less than a development trial between no revenue $LWLG and $TSEM

  11. 质疑 LWLG 估值,拿 Shunsin 和 SIVEF 做对比。

    抱歉,但我越来越确信 $LWLG 就是光子领域里的当前 $RGTI。 我完全不懂……他们怎么会值 18 亿美元。 鸿海的光学业务 Shunsin,给 $NVDA 做 CPO 的封装、组装和测试…… 估值反而只有 14 亿美元? 而 $SIVEF 作为 $JBL 和 $MRVL 的激光源,前几天估值还只有它们的三分之一? 感觉像是傻钱机构因为一份开发测试协议,买错了名字。 我没有持仓,只是觉得困惑。

    英文原文

    Sorry but I’m convinced $LWLG is the current $RGTI of photonics. I have zero clue… How they’re valued at $1.8B. Foxconn’s optical arm Shunsin, for $NVDA CPO packaging, assembly, and testing… Is valued less at $1.4B? $SIVEF, the laser source for $JBL and $MRVL was 1/3rd their valuation earlier this week? Feels like dumb money institutions went into the wrong name off a development test agreement. I have no open positions, just confused

  12. 看好 Shunsin 翻倍,难接受其估值低于 LWLG。

    @TD_btc24 我很看好 Shunsin(6451)翻倍。很难接受鸿海的光学封装业务估值还低于 $LWLG,笑死。

    英文原文

    @TD_btc24 Pretty confident in Shunsin (6451) doubling. Hard to see Foxconn’s optical packaging arm valued less than $LWLG lol

  13. 建议看下一轮光子周期,不必畏惧已涨多的名字。

    如果你对像 $LITE 这种已经涨了很多的名字感到不舒服,你可以去看看下一轮光子周期的曝光。 像鸿海光学封装业务 Shunsin 这种我真的很喜欢……或者 $SIVE 这种 1.6T / CPO 激光,即使在现在这个价位也是。 $SOI 也还有很多空间。 分几天做成本均摊也没坏处。

    英文原文

    If you're uncomfortable with names that already went up a lot like $LITE, you can try for exposure in the next optical cycle. Stuff like Shunsin for Foxconn's optical packaging arm i genuinely really liked... or $SIVE for 1.6T/CPO lasers, even at current levels. $SOI still has a lot of room to go. Doesn't hurt to cost average over multiple days.

  14. 说瑞典本地投资者做空只会火上浇油。

    当地的瑞典投资者都开始做空它了,真是搞笑。 这只会火上浇油,因为他们根本不懂 photonics / hyperscaler 供应链。 不过瑞典媒体倒是在帮西方资金做事……原本是本地散户持有的股票,最后只会被转手给西方机构。

    英文原文

    All the local Swedish investors started shorting it which is hilarious. Just adds fuel to the fire since they have a lack of understanding of photonics/hyperscaler supply chains. Swedish media is doing wonders for Western funds though… it was a local retail owned stock but it’s just going to get transferred to Western institutions later.

  15. 说 SIVE 正在走向 LITE 路径,且有多重利好。

    $SIVE 正在走向变成 $LITE 的路上。 过去两天里: > $JBL 确认使用 $SIVE 激光用于他们的 1.6T 光学收发器 > Sivers 从长期机构投资者(比如养老基金 / 机构投资者)那里融资 1350 万美元以上,降低了资产负债表风险 > $SIVE 计划在美国 NASDAQ 上市。 这走势也太疯狂了。 我还是觉得短期内它会冲到 20 亿美元以上市值。 尤其是随着股票从瑞典本地投资者转向真正懂 photonics 的西方基金 / 投资者。

    英文原文

    $SIVE is on its journey to be $LITE . In the last 2 days: > $JBL confirmation of using $SIVE lasers for their 1.6T optical transceivers > Sivers raises $13.5M+ from long term institutional investors (eg. pension funds/institutional investors), de-risking balance sheet issues > $SIVE plans to list on US NASDAQ. What a crazy turn of events? I still think it's heading to $2B+ MC in the near term. Especially as shares transfer from local Swedish investors to Western funds/investors who understand what’s up and coming with photonics.

  16. 解释 Fittech 与 Shunsin、Win/Sivers 的供应链去向。

    @KentTian Fittech 在 $TSM Coupe 生态里,Shunsin 是鸿海的光学封装。 Win / Sivers 是 $JBL、$MRVL,还有 Ayar(也许还有 AlChip 等)。 它们最终都会流向 hyperscaler,只是路径和终点不同。

    英文原文

    @KentTian Fittech is in $TSM Coupe ecosystem, Shunsin is Foxconn optical packaging. Win/Sivers is $JBL, $MRVL, and Ayar (maybe AlChip and others). They all end up to hyperscalers, just in different ways and different ones.

  17. 看好 SIVE 与 Win Semi 受益于激光量产周期。

    这正是我做多 $SIVE 激光生态加 Win Semi(3105)的原因。 下一轮激光大规模量产的架构超级周期,受益者是谁? Sivers + Win。 尤其是昨天宣布 Sivers 正在为 $JBL 的 1.6T 光学收发器供货之后。 https://t.co/05wYF3fR0U

    英文原文

    This is exactly why I long the $SIVE laser ecosystem with Win Semi (3105). The beneficiaries of the next architectural supercycle for laser mass production? Sivers + Win. Especially with the announcement yesterday that Sivers is powering $JBL 1.6T optical transceivers. https://t.co/05wYF3fR0U

  18. 强调 SIVE 今天有融资和纳斯达克上市双重利好。

    @PolarizingLit 很多人忽略了今天 $SIVE 的双重公告: 1. 他们从长期机构投资者那里融资了大约 1300 万美元以上,所以不再有资产负债表担忧。 2. 他们还在推进美国 NASDAQ 上市。 50% 的跳空上涨其实低估了这条消息。

    英文原文

    @PolarizingLit People miss that it was a DUAL announcement too today with $SIVE: 1. They raised ~$13M+ from long term institutional investors (so no more balance sheet concerns). 2. Then they're pursuing US NASDAQ listing too. 50% gap up is undervaluing the news.

  19. 说 SIVE 融资很小但极其利好,且大概率用于上市。

    是的,$SIVE 从一批机构投资者那里融资了约 1365 万美元,这非常利多。对流通盘来说占比也很小,大约只有 2.5%。 而且这笔钱大概率是为了 NASDAQ 上市(极度正面)。 极度正面 + 极度正面 = 对管理层执行力的高信心。

    英文原文

    Yeah $SIVE raised ~$13.65M from a bunch of institutional investors, which is strongly bullish. For such a tiny portion of the float ~2.5% too. And it's likely used for NASDAQ listing (extremely positive). Extremely positive + Extremely positive = high conviction in management execution.

  20. 强调 SIVE 正在推进纳斯达克上市,像下一只 LITE。

    $SIVE 正在推进美国 NASDAQ 上市,路透已经报了。 欢迎来到美国,Sivers。 我之前说过,基于今天的基本面,Sivers 现在看起来很快就应该值 20 亿美元以上。 而且一旦在 NASDAQ 上市,美国机构很可能会开始涌入这家给 $MRVL 和 $JBL 供货的激光供应商。 这发生在今天一轮非常利好的 2.5% 融资之后,融资对象是新的“国际机构投资者”,大概率是为了支付上市所需的监管 / 审计成本。 “参与定向配股的投资者包括少数瑞典和国际机构投资者及其他合格投资者。” 我们大概率会看到消息推动的近乎抛物线式上涨。 仅仅过去三天: -> $JBL 1.6T 使用 $SIVE -> $SIVE 新的美国 / 国际机构投资者 -> $SIVE 看起来要在 NASDAQ 上市。 我们正在见证下一只 $LITE 的诞生。

    英文原文

    $SIVE executing US NASDAQ listing from Reuters. Welcome to America Sivers. I said Sivers looks like it should be valued at $2B+ soon based on today's fundamentals. And US institutions will likely be piling into the laser supplier for $MRVL and $JBL once it's on NASDAQ. This comes after a highly bullish 2.5% raise today from new "international institutional investors" to likely fund regulatory/audit requirements for the listing. "The investors in the Directed Share Issue comprise of a limited number of Swedish and international institutional and other qualified investors" We’ll likely going to see a parabolic ride up from the news. In just the last three days: -> $JBL 1.6T to use $SIVE -> $SIVE new US/international institutional investors -> $SIVE looks to be listed on NASDAQ. We're witnessing the birth of the next $LITE.

  21. 说 ALRIB 是量子链条里最喜欢的名字。

    $ALRIB 是我在这个赛道里最喜欢的名字,尤其是在市场发现 $MSFT Quantum 会从他们那里采购之后。 现在还稍微有点早,但做机器设备销售的公司会受益于今天整个研发 capex 周期,而且你不需要在 $IONQ、$RGTI、$QBTS、$GOOGL 或其他赢家之间二选一。

    英文原文

    $ALRIB is my favorite for that sector, especially now that markets found out $MSFT Quantum buys from them. It's still a bit early, but the machine sellers benefit from the entire R&D capex cycle today and you don't need to pick between $IONQ, $RGTI, $QBTS, $GOOGL or others winning.

  22. 强调供应链映射前瞻性远胜基本面回看。

    不,他们的基本面并不一样。这就是大多数散户不理解、也最让人沮丧的地方。 $SIVE 真的拿到了 $JBL 这个 hyperscaler 客户,O-Net 也在服务亚洲 hyperscaler。 然后 $MRVL 从 $NVDA 拿到了 20 亿美元,还给出了 Celestial CPO 放量的指引(这会传导到 $SIVE)。 $POET 也给了指引,并表示下半年会开始放量交付,比预期更早。 ASE 甚至直接预计 2026 年下半年 CPO 放量,远早于预期。 所以如果我们把激光环节的未来盈利增长往前映射,它应该会加速很多。 市场定价的是未来增长,不是回头看 2025 年的基本面。

    英文原文

    No, they didn't have the same fundamentals. This is what majority of retail investors don't understand and it's so frustrating. $SIVE literally landed $JBL as a hyperscaler client, and O-Net which serves Asian hyperscalers. Then $MRVL ended up getting $2B from $NVDA and gave projections on their Celestial CPO ramp (which trickles down to $SIVE). $POET also gave projections and volume ramp delivery H2, which was earlier than expected. ASE literally went out and projected CPO volume ramp H2 2026, way earlier than expected. So if we map forward earnings growth from lasers, it should accelerate a lot. Markets price in future growth, not backward 2025 fundamentals.

  23. 认为 Enplas 仍有双重瓶颈,值得在中小市值里关注。

    @Letterkees 我还是认为 Enplas 会在 6 个月内翻倍。 通常你不会看到服务 hyperscaler 的 <10 亿美元公司长期还停留在那个市值区间,而且它们在 photonics / ASIC-GPU 测试上握着两个不同的瓶颈。

    英文原文

    @Letterkees I still think Enplas doubles within 6 months. Normally you don't see &lt;$1B companies serving to hyperscalers stay in that MC range for too long and they own two different chokepoints in photonics/ASIC-GPU testing.

  24. Sivers成为JBL光学收发器激光光源供应商,证实进入超大规模业者供应链。

    Sivers Photonics ($SIVE) 是下一个 $LITE,市值5.6亿美元。机构投资者今天得到了完整确认:Sivers 现已成为超大规模业者(hyperscaler)供应链中的光源(light source),并直接向 $JBL 供应光学收发器(optical transceivers)。这只是时间问题。 官方消息:$JBL 将使用 $SIVE 的激光器用于光学收发器。今天的消息:"Jabil 计划开发一款1.6T线性接收光学(LRO)收发器模块,使用 Sivers 的高性能分布反馈(DFB)激光器"。Jabil Photonics 表示:"与 Sivers 合作将使我们能够提供满足数据中心性能及功耗目标的大规模1.6T LRO解决方案"。 你是否见过哪个光源(light source)进入超大规模业者供应链...市值只有5亿美元?我们此前在 OFC 展会上从物理渠道获得过暗示,但许多机构投资者需要类似这样的实际确认。

    英文原文

    $SIVE is the next $LITE at $560m MC. Institutions just got full confirmation today: Sivers is now the light source in hyperscaler supply chains and the direct supplier of $JBL optical transceivers. It’s only a matter of time. https://t.co/Eh9GherJTV

  25. 强调 JBL 的规模会放大 SIVE 的影响。

    @StormDirac 没错。$JBL 体量巨大,尤其是在它更早收购了 $INTC 的 SiPH 部门之后。 我真的不觉得市场已经意识到 $SIVE 接下来会发生什么。

    英文原文

    @StormDirac Exactly. $JBL is massive, especially after buying $INTC's SiPH division earlier. I genuinely don't think markets realize coming yet with $SIVE.

  26. 把 SIVE 与 JBL 的供货关系视作结构性转变。

    这终于算是 $SIVE 的供应链映射验证了。 $JBL 真的******巨大,而 Sivers 激光现在大概率会进入 Jabil 的 hyperscaler 客户,比如 $AMZN、$GOOGL。 这是一个非常结构性的变化,如果官方确认下来,我个人会把 Sivers 估到 20 亿美元以上。

    英文原文

    This is finally supply chain mapping validation with $SIVE. $JBL is ****** massive and Sivers lasers now will likely go into Jabil's hyperscalers clients like $AMZN, $GOOGL. This is such a structural change and I would personally value Sivers over $2B+ after official confirmation.

  27. 官方确认 JBL 将使用 SIVE 激光。

    官方确认:$JBL 将为他们的光学收发器使用 $SIVE 激光。 今天的消息: “Jabil 计划使用 Sivers 的高性能分布反馈(DFB)激光,开发 1.6T 线性接收光学(LRO)收发器模块。” Jabil Photonics: “与 Sivers 合作将使我们能够大规模交付满足数据中心性能和功耗目标的 1.6T LRO 方案。” 你在哪儿见过 hyperscaler 供应链里的‘光源’…… 而且还是在 5 亿美元市值下? 我们之前在 OFC 物理渠道里已经听到过暗示,但很多机构需要这样的正式确认。

    英文原文

    IT'S OFFICIAL: $JBL to use $SIVE Lasers for their optical transceivers. Today: "Jabil plans to develop a 1.6T linear receive optical (LRO) transceiver module using Sivers’ high-performance Distributed Feedback (DFB) lasers" Jabil Photonics: :Working with Sivers will allow us to deliver a 1.6T LRO solution that meets both data center performance and power targets at scale" Where have you seen the LIGHT SOURCE for hyperscaler supply chains... At a $500m MC? We had this hinted from physical sources at OFC, but many institutions needed actual confirmation like this.

  28. 指出 Shunsin 就是 Nvidia CPO 供应链里的鸿海光学封装。

    @KDREAM111111 你是不是傻。Shunsin 就是鸿海的光学封装业务,处在 $NVDA 的 CPO 供应链里,而且 forward P/E 很低。

    英文原文

    @KDREAM111111 Are you stupid. Shunsin is literally Foxconn's optical packaging arm with low forward P/E in $NVDA's CPO supply chains.

  29. 认为 Enplas 的双瓶颈与客户结构很有意思。

    Enplas(6961)在大约 9.85 亿美元市值时看起来挺有意思。 它们握着两个瓶颈: 1. MLA(微透镜阵列)的主导供应商。比如 SiPH switch、1.6T、3.2T 光子产品。 2. IC 测试 socket 的寡头供应商(AI 芯片测试) 手头现金大约 1.55 亿美元,没有债务。股东权益/资产比大约 89%(一般 50%+ 就算稳),所以我觉得下行风险很低。 而且它们推测的客户是: 1. 光子(MLA):高度可能是 Innolight、Eoptolink、Furukawa、Intel(SiPh)。大概率还有 $COHR、$LITE。 2. GPU/ASIC(测试 socket):大概率是 $TSM、ASE,以及这些类型用来测试 $NVDA GPU、$GOOGL TPU(Google Ironwood 很可能就是这个方向)。 “为主要 GPU 制造商以及 hyperscaler 的 ASIC 相关项目扩大量产订单。” 然后在 OFC: OFC: “我们将展示支持当今 AI 快速技术进步的 800Gbps 和 1.6Tbps 收发器以及 CPO(共封装光学)的新产品。” 本质上你得到的是一家同时服务一线半导体、晶圆厂和 hyperscaler 的公司,并且它能同时受益于 MLA 光子段带来的 1.6T,以及之后 OFC 产品带来的 CPO TAM 扩张。 这是我想分享的一笔持仓,因为它在不到 10 亿美元市值时就显得挺酷。

    英文原文

    Enplas (6961) at ~$985M seems kinda interesting. 
They hold two chokepoints: 1. Dominant supplier for MLAs (micro lens arrays). Eg. SiPH switches, 1.6T, 3.2T for photonics. 
 2. Oligopoly supplier for IC Test sockets (AI Chip Testing) Cash on hand: ~$155M, no debt. Equity-to-Asset Ratio: ~89% (50%+ usually is solid), so low downside risk imo. 
And their speculated customers: 1, Photonics (MLAs): Highly probable: Innolight, Eoptolink, Furukawa, Intel (SiPh). And prob $COHR, $LITE. 2. GPU/ASICs (Test Sockets): prob $TSM, ASE, types use these to likely test $NVDA GPUs, $GOOGL TPUs (Google Ironwood is highly probable). "expanding mass production orders for major GPU manufacturers, and for ASIC-related projects for hyperscalers."
 Then at OFC: OFC: “We will be showcasing our new products for 800Gbps and 1.6Tbps transceivers and CPO (Co-Packaged Optics) that support today’s rapid technological advancements in AI.” Basically you have a company that supplies T1 semis, foundries, hyperscalers, that benefits from 1.6T from MLA photonics segment + CPO TAM expansion later from OFC products. Was one my positions wanted to share my thoughts about since it seemed p cool at sub <$1B MC.

  30. 指出 Nippon Chemical 是 photonics AI 供应链关键材料商。

    @Jornka329996 哦,那家是 Nippon Chemical Industries,市值 1.69 亿美元;如果把它打掉,整个 photonics 方向的 AI buildout 都会被扰乱。 因为它是生产 InP 衬底所需红磷的第一供应商,而这些衬底又要用在 $AXTI、Sumitomo 和 JX 里。

    英文原文

    @Jornka329996 Oh it was Nippon Chemical Industries at $169m, if you take down that company it disrupts the entire AI buildout with photonics. Since its the #1 supplier of red phosphorus needed to make inp substrates at $AXTI, Sumitomo, and JX.

  31. 列出硅光晶圆测试瓶颈相关的几个标的。

    “硅光扩产碰到了晶圆测试瓶颈” - $FORM(ficontec 离开后更有帮助),$AEHR - 台湾这边有 MPI(6223)、MA-tek(3587)、MSScorps(6830) - 再往上还有大市值的 $TER、Advantest(6857) 如果你感兴趣,这些大概是我脑子里能立刻想到的一些不错曝光。 我很快会给大家做一个更深入的 ETF 版本。

    英文原文

    "Silicon photonics scaling hits wafer testing bottleneck" - $FORM (helps that ficontec left), $AEHR - Over in Taiwan MPI (6223), MA-tek (3587), MSScorps (6830) - Then large cap Cap - $TER, Advantest (6857) Are some decent exposure off the top of my head if you’re curious. I’ll make a deeper dive ETF for you all soon.

  32. Soitec 反弹后继续看好 CPO/硅光主线。

    自从我发出 TLDR 版 thesis 后,Soitec 现在一个月涨了 64%。 很多人在我发帖后又因为伊朗 / 宏观回调,把 $SOI 说成“没有基本面的 meme 股”。 但我的多头,尤其是像 $SIVE 这类会延伸到 2029 年的 CPO / Silicon Photonics 相关标的…… 有时候就是需要时间兑现? 它在衬底层面是真正的垄断,而且它那些旧业务拖累板块也很可能正在见底。

    英文原文

    Soitec is now up 64% in 1 month since my TLDR thesis post. Lot of people made fun of $SOI as a “no fundamentals meme stock” when it corrected after I posted (because of Iran/macro). But my longs, especially with CPO/Silicon Photonics related stuff like $SIVE that scales into 2029. Sometimes just need time to play out? It’s a genuine monopoly in the substrate layer and it’s likely bottoming in their legacy drag segments.

  33. 越看越觉得 ALRIB 很有吸引力。

    $ALRIB 这只我开始越看越顺眼了。 首先,它有 $MSFT Quantum 作为核心 hyperscaler 客户。 然后它还有像 QD Laser 这样的量子点客户……以及像 $IQE 这样的光子晶圆代工客户。 而且它是盈利的,还在分红…… 在 MBE 领域也是真正的双寡头。 再加上新的硅光机器突破,订单簿还在加速增长…… 我一直把 $LWLG 和 Riber 这种东西拿来对比,然后就一直在想: 这在当前价格下,难道不就是很有吸引力吗?

    英文原文

    $ALRIB is starting to grow on me. First they have $MSFT Quantum as their core hyperscaler customer. Then they have Quantum Dot like QD Laser… and photonics foundries like $IQE as customers. And it’s profitable + paying dividends… And a genuine duopoly in the MBE space. And accelerating order book growth with a new breakthrough in silicon photonics machines.. I keep comparing $LWLG with stuff like Riber, then I keep thinking: Isn’t this just highly compelling at current prices?

  34. 认为市场只是发现了一个 AI 基础设施脆弱点。

    @beauty_oe 没办法啊!这只是单纯的信息发现和整合,市场自己后来才意识到而已。 我发帖的时候它还按 0.5 倍市净率交易,所以就算涨了 80%,也只是回到 1 倍市净率而已…… 而且那还是整个 AI 基础设施构建链条里的一个重大脆弱点。

    英文原文

    @beauty_oe 仕方ないじゃん! 単なる情報の発見と統合で、市場がそれに気づいただけだよ。 私が投稿した時はPBR0.5倍で取引されてたから、たとえ80%上がったってやっとPBR1倍に戻るだけだし… しかもそれって、AIインフラ構築全体における重大な脆弱性でもあったりするんだよね。

  35. 说自己最早把 AXTI 和 photonics 瓶颈联系起来。

    @Kelvinccccccccc 也许是因为没人知道吧? 我是最早把 $AXTI 和 photonics 瓶颈联系起来的人之一。 而 NCI 是更小众、位于上游的红磷供应商,而这些红磷又是制造 InP 衬底所必需的。 所以它其实是“瓶颈里的更瓶颈”的那种存在。

    英文原文

    @Kelvinccccccccc Maybe because nobody knew about it? I was one of the first people to point out $AXTI in relation to photonics bottlenecks. And NCI is an even more the niche upstream red phosphorus supplier needed to make InP substrates. So the even more unknown bottleneck of a bottleneck.

  36. 补充 NCI 才是高纯红磷的主导供应商。

    rasa 虽然也是双寡头之一,但和 NCI 比,它在红磷这块更多算是次级供应商。 我更在意的是集中度比例,而 NCI 是全球领头羊。不过没错,你可以买两家,而西方世界基本上已经没法自己生产高纯度红磷了…… 也就没法再做 InP 衬底……也就没法再做 photonics。

    英文原文

    rasa is a duopoly but they're considered a secondary supplier for red phosphorus compared to NCI. I cared more about concentration %, where NCI is the world's leader. But yes, you can buy both and the Western world basically can't make high purity red phosphorous anymore... and no more inp substrates... and no more photonics.

  37. 说日本出口管制能反制中国的红磷供应。

    $AXTI 确实依赖像 NCI 这样的公司,所以中国的 AI 供应链实际上会被日本卡住。 所以我希望日本能在高纯红磷上用出口管制把中国反制回去,算是对 Jan 当年出口管制的回敬。 不过中国也在提升本土产能,但目前还不够。

    英文原文

    $AXTI does rely on companies like NCI, so China's AI supply chains would actually get bottlenecked by Japan. So I hope Japan export controls China back on high purity red phosphorus as payback for Jan's export controls. China is scaling up domestic capacity though but it's still not enough yet.

  38. 用 1.69 亿美元撬动西方 hyperscaler 供应链。

    Serenity 指南: 如何只用 1.7 亿美元就让西方 hyperscaler buildout 变得更脆弱。 只要用 1.69 亿美元拿下 Nippon Chemical(4092)! 做 InP 衬底需要:铟和高纯磷。 你以为 $AXTI 是瓶颈?NCI 才是瓶颈里的瓶颈。 NCI 实际上是高纯红磷卡点的领头羊,份额大约 26-27%(Rasa 更少,其余就是中国)。 而且它们向 $AXTI、Sumitomo、JX 供应这些材料,用来做 InP 衬底。 所以……如果你拿出 1.6 亿美元收购 NCI(再加上 Rasa 作为更小产能),你就能移除西方世界生产 6N/7N 红磷、也就是制造 InP 衬底所需的主力产能! 没有 InP 衬底:就没有 photonics。 有趣的是:中国科技公司其实也会被 NCI 打乱。 对 $AXTI 来说,这个映射 / 依赖关系其实非常有意思: - AXT 的 Tongmei 在 STAR Market 上市时就披露过,它在结构上依赖从日本进口高纯前驱体材料 - WITS 数据显示,大约 460 美元 / 公斤的高纯磷正从日本流向中国 所以它们确实在暗中依赖 NCI。 中国虽然也有 Wylton Chemical、Qin Xi New Materials、Jinding Electronics、Chuxiong Chuanzhi、Guizhou Wylton Jinglin Electronic Materials 这类产能。 但这些都还是更小的玩家,没法在规模化 InP 衬底生产上替代 NCI 提供的高纯红磷产能。 $LITE 的 CEO 之前就说过 InP 衬底让他夜不能寐。 所以现在有了 NCI,你就能让他永久失眠? 只要 1.69 亿美元。 所以供应链大概是这样: -> DGC 磷矿开采后运给 NCI -> NCI 把黄磷精炼成高纯红磷 -> Sumitomo / JX / AXT 把红磷和铟熔合,长成 InP 衬底 -> $COHR / $LITE 把 InP 衬底做成激光器 -> Innolight / Fabrinet 把它们封装成 800G / 1.6T 收发器 -> $NVDA / $GOOGL 用在 ASIC / GPU 集群里。 基本上,整个西方都依赖 NCI 来制造用于 photonics 的 InP 衬底。 我持有一点很小的仓位,纯属好玩。不过日本并不以随意提价著称。 所以如果你真的买下这家公司然后把价格抬高 15000%(比如政府意识到后把公司收回去),大概率会碰到监管问题,比如 FEFTA…… 也许提价 30-50% 还能压缩一下 forward P/E?但大概率不会最后变成 $AXTI 那样。 不管怎样,这家公司按大概 1.6 亿美元市值定价,本身就是一个巨大的国家安全风险。 至于基本面,它们市净率 0.54、forward P/E 11.4,所以其实也可能是被低估了。 TLDR: -> 它是下一个 $AXTI 吗?不是。 -> 它是不是西方 hyperscaler buildout + photonics 里一个未知的结构性瓶颈和关键漏洞?是。 -> 还有没有重估空间? 只要回到 1 倍市净率,立刻就有 80-85% 的上涨空间。如果再给它更高的 forward P/E,可能更多。 总之,用 1.69 亿美元找到 hyperscaler 供应链里的一个主要失效点,挺好玩的。

    英文原文

    A Guide by Serenity: How to Cripple the Western Hyperscaler buildout with just $170m. Just take over Nippon Chemical (4092) with $169m! For InP substrates, you need: Indium and High Purity Phosphorus. Thought $AXTI was a bottleneck? NCI is the bottleneck of the bottleneck. NCI is actually the leader of the high purity red phosphorus chokepoint holding 26-27% of the market share (Rasa has less share, then the rest is China). And they export to $AXTI, Sumitomo, JX that need it to make InP substrates. 
So… if you have $160m to spend to acquire NCI (plus Rasa as smaller capacity), you can remove the leading Western world’s production of 6N/7N red phosphorus needed to make InP substrates! And without InP substrates: no photonics. Fun fact: China’s tech companies would get pretty disrupted with it too by NCI. For $AXTI, the mapping/reliance is actually pretty interesting: - AXT's Tongmei outlined its structural reliance on importing high-purity precursor materials from Japan on their STAR Market listing 
- WITS data showing ~$460/kg high-purity phosphorus flowing from Japan into China So they secretly do depend on NCI. 
China does have capacity like Wylton Chemical, Qin Xi New Materials, Jinding Electronics, and Chuxiong Chuanzhi, Guizhou Wylton Jinglin Electronic Materials as well. However, they’re all smaller players so can’t make up for high purity red phosphorus capacity provided by NCI for InP substrate production at scale. $LITE CEO already said inp substrates keeps him up at night. So now with NCI, you can give the guy permanent insomnia? For just $169M. So here's what the supply chain looks like: -> DGC phosphate rock mine and ships it to NCI -> NCI refines Yellow Phosphorus into High Purity Red Phosphorus -> Sumitomo / JX / AXT melt the Red Phosphorus with Indium to grow InP Substrates -> $COHR / $LITE fab InP substrates into Lasers -> Innolight/Fabrinet package them into 800G/1.6T transceivers -> $NVDA / $GOOGL use them for ASIC/GPU clusters. 
And basically, the entire West depends on NCI to make InP substrates for photonics. I hold some very small positions, just for fun. However, Japan is not well known for price hiking. So you’d probably run into regulatory problems eg. FEFTA if you bought the company and hiked prices 15000% (like government seizing back the company once they realize)… Maybe 30-50% hikes is possible to compress fwd p/e? But very likely wont end up like $AXTI. 
 Regardless, this company is a massive, massive national security risk priced at ~$160m. As for fundamentals, they’re trading at .54 book value and a forward P/E of 11.4 so it’s probably undervalued anyway. TLDR: -> Is it the next $AXTI? No. -> Is it an unknown structural bottleneck + critical vulnerability of the Western hyperscaler buildout with photonics? Yes. -> Is there still room for re-rating? Just reverting to Book Value of 1 is immediate 80-85% upside. Maybe more if you give it multiples past 11 fwd p/e. Regardless, it’s fun to find a major point of failure in the hyperscaler supply chains for $169m.

  39. 说 Shunsin 是自己第二喜欢的 NVDA 生态 CPO 标的。

    @dijkstra_levi Shunsin 是我在 $NVDA 生态里第二喜欢的 CPO 标的。

    英文原文

    @dijkstra_levi Shunsin was my #2 favorite $NVDA ecosystem CPO name.

  40. 说美国上市会显著抬升 SIVE 估值。

    @sbloom134 在美国市场 uplisting 应该会让它估值提高不少,因为相当一部分美国机构没法直接买瑞典市场或 OTC。

    英文原文

    @sbloom134 Uplisting in US markets should increase its valuation by quite a bit since a decent amount of US institutions can’t access Swedish markets or OTC.

  41. 认为 SIVE 若是美股会值 20 亿以上,并有清晰上行路径。

    我其实觉得如果 $SIVE 作为美国公司挂牌,今天就应该交易在 20 亿美元以上市值……而不是等到今年晚些时候。 因为我不太明白: - $LWLG 为什么会交易到 3.5 倍以上的估值 - 做自己激光器封装的公司,估值却只有它的 5-6 倍 - 像从 $MTSI 到 $LITE 这些激光公司,估值都在几十亿美元区间。 而 Sivers 又是通过 $MRVL、$JBL、O-Net 等进入 hyperscaler 供应链,而不是只依赖单一客户。 全球公开上市的 AI 数据中心激光公司其实也没多少。 所以它要么会成为 $AVGO 或 $MRVL 很想纵向整合上游的高优先级收购标的。 要么它可以走 $LITE 那种路线:从 17 美元涨到 800 美元,并通过 IP 收购把 ELS / optical transceiver stack 的下游 TAM 扩大。 当然,这取决于公司到底有多激进,但我就是觉得自从我发 thesis 后,没人真正注意到瑞典这个激光供应商。 可能只是需要给美国上市一点耐心,我几乎可以确定美国机构会像对 $IQE 那样感兴趣。 DD 周期通常在他们读完我的 thesis 后会持续几周,但我看得到未来几年从这里走到 100 亿美元以上市值的清晰路径。

    英文原文

    I actually thought $SIVE should be trading at $2B+ MC today (from ~$520m) if they were listed as a US company. Not later this year. Since I’m not sure how: - $LWLG trades at 3.5x+ the valuation - packaging companies that buy their lasers trade at 5-6x their valuation. - laser companies from $MTSI to $LITE have premiums trading in the tens of billions. And Sivers are in hyperscaler supply chains through $MRVL, $JBL, O-Net, and others rather than being dependent on one customer. There’s not very many publicly listed AI DC laser companies in the world either. So either it’s a highly sought after acquisition target for $AVGO or $MRVL that want to vertically integrate upstream. Or they can pull a $LITE that went from $17->$800 and downstream TAM expansion the ELS/optical transceiver stack through IP acquisitions. Depends how ambitious the company is of course but i just don’t think anyone noticed this laser supplier in Sweden since my thesis post. Just might require a little patience for the US listing and I’m almost certain US institutions are interested like they were with $IQE. DD periods usually last a few weeks after they read my thesis but I see a clear path to $10B+ MC from here over the next few years.

  42. 说日本用户无法交易 6451 很可惜。

    @beauty_oe 6451.,日本人居然买不了(交易不了)真让人震惊。那本来可是我个人非常看好的票! 在 $NVDA 的 CPO 供应链里,像那样已经高度去风险的公司可不多。

    英文原文

    @beauty_oe 日本の人たちが6451にアクセス(取引)できないのはショックだな。個人的に銘柄としてすごく推してたのに! $NVDA のCPOサプライチェーンで、あそこまでディリスクされた企業はレアだったよ。

  43. 认为鸿海光学封装业务估值太低。

    @veggiepoultry 是啊,确实如此。真不敢相信鸿海的光学封装业务市值只有 14 亿美元,笑死。 你等于从母公司白拿了一程。

    英文原文

    @veggiepoultry It is. Still can't believe Foxconn's photonic packing arm is valued at $1.4B MC lol. You get a free ride from the parent company.

  44. 说 FOCI 可能排在后面但仍很靠前。

    @MindQuest42 @AstuteKara 不,FOCI 大概排得更后面一点,但仍然算很靠前。

    英文原文

    @MindQuest42 @AstuteKara Nope, FOCI is probably further down the list but still high up there.

  45. 说自己更喜欢 Shunsin,排名第二。

    @AstuteKara 对,我个人更喜欢 Shunsin。它大概是我现在最喜欢的 $NVDA CPO 供应链标的里的第二名。 第一名那只因为涨太多被我冻结了。

    英文原文

    @AstuteKara Yeah I personally preferred Shunsin. It's probably my #2 favorite $NVDA CPO supply chain pick right now. My #1 is frozen because it went up too much.

  46. Shunsin 因为涨太多被停牌。

    看起来我选的 $NVDA CPO 供应链标的 Shunsin…… 鸿海的光学子公司因为涨太多被停牌了。 有趣的是:台湾股票一天最多只能涨 10%。 https://t.co/6dOGew2aGY

    英文原文

    Looks like my $NVDA CPO supply chain pick Shunsin… Foxconn’s optical subsidary got halted because it went up too much. Fun fact: TW stocks can only go up 10% a day. https://t.co/6dOGew2aGY

  47. 用台币价位说明目标价。

    @PatientGS 2240,按 440 台币算,12 个月。

    英文原文

    @PatientGS 2240 from 440 TWD, 12 months

  48. 说 LITE 已卖到 2029 年,CPO 量产已确认。

    @itsakek - $LITE 已经卖到 2029 年了。 - 2026 年下半年确认 CPO 量产。 $NVDA CPO / photonics 上游供应链直接起飞。

    英文原文

    @itsakek - $LITE sold out into 2029. - Confirmation CPO mass production H2 2026. $NVDA CPO/photonics upstream supply chain go brrr

  49. 给出 2027 年业绩模型和 EPS 推演。

    好问题!大致是这样: - 2027 年营收:新台币 127.2 亿 - 毛利(26% 毛利率):新台币 33 亿 - 运营费用(约新台币 11 亿) - 营业利润:新台币 22 亿 - 综合税费 + 非经营性:新台币 3.3 亿 - 净利润:新台币 18.7 亿 - 每股收益约新台币 17.61+,向 20+ 看 前些年因为在河内 / 北江扩建 CPO 设施,资本开支和折旧都很多,所以毛利率扩张、产能利用率到 80% 后会去到 26%。 模型基于 Shunsin 拿下 350-400 万个 SiPh/CPO 封装单元,ASP 为新台币 2,350。 350 万 × 2350 = 新台币 82 亿 传统业务(稳定后):新台币 45 亿 8.2 + 4.5 = 2027 年营收 127 亿。并且还应该能随着 2028、2029、2030 年的光学预期继续扩张。

    英文原文

    Good question! Basically: -NT $12.72B revenue 2027 -Gross profit (26% margin): NT $3.3B -Opex (~NT$1.1B) -Operating income: NT $2.2B -Blended Tax + non-op: NT$330m -Net income: NT$1.87B ~NT$17.61+ EPS to $20+ Lot of capex/depreciation in former years because of scaling up CPO facilities in Hanoi/Bac Giang, so gross margin expansion and 80% utilization to 26%. Modeled based off Shunsin capturing 3.5-4M SiPh/CPO packaging units witH ASP NT$2,350 3.5M × NT2,350= NT $8.2B Legacy Business (Stabilized): NT$4.5B 8.2 + 4.5 = 12.7B rev for 2027. And should scale up based on optical projections for 2028, 2029, 2030.

  50. 说 6451 像是鸿海光学业务的低估去风险化标的。

    @latent_value7 对……我觉得这就像一趟低市值、低 forward multiple 的鸿海光学业务去风险化搭便车。 我真的很喜欢 $6451.TW。

    英文原文

    @latent_value7 Yeah... looks like a de-risked foxconn optical segment piggy back ride to me at a low MC and low fwd multiples. I really liked $6451.TW.