NVDA英伟达
27看多0看空93提及
Serenity @aleabitoreddit
专注 AI 光通信与半导体供应链研究，擅长从产业链与公司基本面寻找高赔率机会。长期跟踪 CPO 与光模块产业链，持续输出 SIVE、AAOI、LITE、POET 等标的的深度观点。
结论速览
NVDA 整体偏多（看多 27 · 看空 0）
被提及为客户，其自动驾驶架构标准由AEVA提供，而AEVA由SIVE供货。
最近发言：2026/08/24 17:05 · 信心分 9/10
8-10 分观点使用大徽标▲ 看多▼ 看空● 提及
历史观点（120 条，最新在上）
- English · 原文Not exactly, I think $AAOI stated they avoided LTAs so they wouldn't have too much capacity committed to one customer and have to turn away others. If they signed LTAs with prepayments, then that in scenario there probably wouldn't need to be such sizeable ATMs. Decision was a bit surprising though considering how well $NVDA LTAs turned out well for $LITE / $COHR. I would assume though AOI gets more pricing leverage as time goes on due to current laser/transceiver shortages.中文 · 翻译不完全是，我觉得 $AAOI 表示他们之所以避免签订长期供货协议（LTA），是为了不把太多产能绑定给一个客户，导致不得不拒绝其他客户。 如果他们签订了附带预付款的 LTA，那在这种情况下，可能就不需要那么大规模的 ATM 增发了。 不过这个决定还是有点出乎意料，考虑到 $NVDA 的 LTA 对 $LITE / $COHR 来说效果都很好。 不过我猜，由于目前激光器/收发器短缺，AOI 随着时间推移会获得更多定价权。
- English · 原文$NVDA, following their $20B deal... Announces Groq3 LPX now in full production. With $NBIS the first to adopt through Nebius token factory. - 3,400 tok/s in Artificial Analysis benchmarking running Gemma 4 31B, with 100k token context - 4x faster responsiveness than the nearest alternative Thing that caught my attention.. was a former press release after Nvidia's $2B investment into Nebius. That it would support their "early adoption of Nvidia's latest generation" architectures. Guess we're seeing that special treatment here.中文 · 翻译$NVDA，继他们那笔200亿美元的交易之后……宣布 Groq3 LPX 已全面投产。 $NBIS 是首个通过 Nebius 代币工厂采纳这项技术的。 - 在 Artificial Analysis 的基准测试中，运行 Gemma 4 31B，每秒3,400个token，支持10万token上下文 - 响应速度比最接近的替代品快4倍 让我注意的是……英伟达向 Nebius 投资20亿美元后发布的一份新闻稿。 里面提到会支持他们“早期采用英伟达最新一代”的架构。 看来我们现在看到的就是这种特殊待遇了。
- English · 原文$XPEV carves out its robotics unit and raises >$900M at $6.3B+ valuations. This is following’s Unitree IPO, which now has a ~$35-$36B MC. Then there’s $NVDA / $AMZN backed Agility… at a $2.5B premoney valuation. Which I’d argue is actually further ahead in commercialization than Xpeng. Regardless, humanoids + physical AI sector funding is really taking off.中文 · 翻译$XPEV 剥离了机器人部门，并以超过 63 亿美元的估值融资了 9 亿多美元。 这是在 Unitree IPO 之后发生的，Unitree 现在市值约为 350-360 亿美元。 还有 $NVDA / $AMZN 支持的 Agility… 投前估值 25 亿美元。我个人认为，它在商业化方面实际上比小鹏更领先。 不管怎样，人形机器人 + 实体 AI 领域的资金投入真的爆发了。
- English · 原文$MRVL gives $GOOGL options to buy $12.2B of the company. "The Marvell-Google deal covers a broad range of technologies used with TPUs, including processors that run AI models, manage data storage and move information across networks." Which could translate into ~$120 billion in revenue through 2033. Marvell also has separate warrants with $AMZN after their Celestial acquisition (for purchases of photonic fabric). Remember the good times when Jensen said Marvell would be the next $1T+ company? Seems Nvidia knew Marvell was "networking" its way into all the hyperscalers with warrants.中文 · 翻译$MRVL 授予 $GOOGL 期权，可以购买价值 122 亿美元的公司股份。 "Marvell 和 Google 的这笔交易涵盖了广泛的技术，用于 TPU，包括运行 AI 模型、管理数据存储和在网络中传输信息的处理器。" 这有望在 2033 年前带来约 1200 亿美元的
- English · 原文Unitree (688836) is now public. And is valued at ~$53.3B, after rising +492.18% We finally have a major public company benchmark for humanoids. For reference, Agility (backed by $NVDA, $AMZN) via $CCXI is expected to go public at $2.5B premoney valuation Q4. $TSLA is $1T+, but Optimus is wrapped inside a much larger company. But I think Unitree showed that the demand for pure play humanoid players is much larger… than people expected in public markets.中文 · 翻译Unitree (688836) 现在上市了。 并且在上涨 492.18% 之后，估值达到了约 533 亿美元。 我们终于有了一个人形机器人领域的重大上市公司基准。 作为参考，Agility (由 $NVDA、$AMZN 支持) 通过 $CCXI 预计将在第四季度以 25 亿美元的投前估值上市。 $TSLA 市值超 1 万亿美元，但 Optimus 被包裹在一家大得多的公司里面。 但我认为 Unitree 表明，公开市场对纯粹的人形机器人公司的需求远比人们预期的要大得多……
- English · 原文My $AAOI TLDR notes at Rosenblatt summit: - Expected to get paid premiums for US production of 800g/1.6T (very positive for ASP/margins). - Several LTAs on the table, but doesn't want to sign to get capacity blocked by other customers (cough cough $NVDA). - Sold out at least through second half of next year and beyond. (High demand visibility like $LITE) - Has 300-400 mW lasers already. - Expects margins to be above 40%+ once CPO comes about (probably most material for rerating). I can't see how anyone can be bearish on this company...中文 · 翻译我在 Rosenblatt 峰会上的 $AAOI 简报要点： - 预计美国生产的 800g/1.6T 产品能获得溢价（这对平均销售价格/利润率非常有利）。 - 有好几个长期供货协议（LTA）摆在眼前，但不想签，怕产能被其他客户占用（咳咳，$NVDA）。 - 订单已排满，至少到明年下半年甚至更久。（需求可见度很高，像 $LITE 一样） - 已经有 300-400 毫瓦的激光器了。 - 预计一旦 CPO 普及，利润率能超过 40%+（这可能是公司价值重估最重要的因素）。 真搞不懂谁会看空这家公司...
- English · 原文I still have positions in both companies? What I hold shouldn't affect your own decisions. Transformer bottlenecks are stalling the US AI buildout now and lead times are very long for both dry and liquid (which is more extreme). Backlog is +96.9% Y/Y for $HPS.A and given their market share... I don't they're going anywhere. Pretty sure I shared this idea at ~170, it did 2x, but corrected back to 240 now, but cost average is different per person I guess. The Belgian company is primarily H1 2028 as a CPO/photonics volume ramp beneficary. And the work they're doing with MTP with IMEC for $NVDA as an end user in evaluation stage should be very interesting. Their CHIP act facility for MEMS/photonics also comes online H2 2028 too, and 800V transitions should bring them material revenue growth for anything SiC adjacent next year. I'm down -31%, since my cost average is a lot higher. But they're approaching levels where replacement book value is a lot higher than MC.中文 · 翻译我在这两家公司都还有持仓？我的持仓情况不应该影响你的决策。 现在，变压器瓶颈正在阻碍美国的AI建设，而且无论是干式还是液式（液式更极端），交货周期都非常长。 $HPS.A 的积压订单同比增长了96.9%，再考虑到他们的市场份额…… 我不觉得他们会出什么问题。 我很确定我大概在$170的时候分享过这个想法，它涨了两倍，但现在又回调到了$240，不过每个人的成本价肯定不一样。 这家比利时公司主要会在2028年上半年受益于CPO/光子学产能的大规模增长。而且他们正在与IMEC合作，为作为终端用户的 $NVDA 进行MTP的工作，目前处于评估阶段，这应该会非常有趣。 他们根据《芯片法案》建造的MEMS/光子学工厂也会在2028年下半年投入运营，明年，800V电压的转型应该会给他们在任何与SiC相关的业务上带来显著的营收增长。 我亏了31%，因为我的平均成本要高得多。但它们正在接近一个水平，在这个水平上，重置账面价值远高于市值。
- English · 原文@tRbW6qAUT717891 I can't buy A-Shares, so I'm sticking with US markets. Agility via $CCXI( $NVDA, $AMZN, Softbank backed) is my personal exposure to humanoids. But I'm cheering on everyone who won the lottery for Unitree shares.中文 · 翻译@tRbW6qAUT717891 我买不了A股，所以还是专注美股吧。 通过 $CCXI ( $NVDA、$AMZN 和软银投资的) 投资 Agility，是我个人投资人形机器人赛道的途径。 但我还是为所有中签宇树科技股票的朋友们欢呼。
- English · 原文I feel like AI investing is simpler than what people expect. Because $NVDA + Jensen literally tells everyone what's coming. But somehow. Almost every. single. time. Markets dismiss it until it actually happens? Be Nvidia in 2025: Buys up EML and laser capacity. Markets dismissing it: "Photonics is a bubble and like quantum! ____ company is a scam with shady management" 1 year later: $LITE +678%, $AAOI +475.12%, $COHR +260.7%, $AXTI +3,843.9%. _ Nvidia in 2026: Buys up CW/EML capacity with LTAs. 800V shift. Extraordinary explicit about CPO shift. States Physical AI as the next theme. Markets now: "CW players are meme stocks! 800v, CPO is not coming anytime soon, Humanoids are not profitable!" Yeah... We'll see what happens in 2027. I think I'm putting my money on Jensen/Nvidia as the leading indicator.中文 · 翻译我感觉AI投资比大家想象的要简单。 因为$NVDA和黄仁勋（Jensen）简直就是把未来会发生什么明明白白地告诉所有人。 但不知怎么的。几乎每一次。每一次！市场都会不屑一顾，直到事情真正发生？ 想象一下2025年的英伟达：收购了EML和激光产能。 市场不屑一顾：“光子学是个泡沫，就像量子技术一样！____公司管理层有问题，就是个骗局！” 一年后：$LITE 涨了678%，$AAOI 涨了475.12%，$COHR 涨了260.7%，$AXTI 涨了3,843.9%。 _ 2026年的英伟达： 通过长期供货协议（LTAs）收购了连续波（CW）/EML产能。转向800V架构。非常明确地提到了CPO（共封装光学）的转变。明确指出实体AI是下一个主题。 现在的市场：“CW公司都是meme股！800V、CPO短期内不会实现，人形机器人不赚钱！” 是啊……我们拭目以待2027年发生什么吧。 我想我还是把钱押在黄仁勋/英伟达这个领先指标上。
- English · 原文I'm not sure if it's just me, but I'm just seeing enormous CW demand imbalance right now (thank you $NVDA). And this is before scale up and other 1.6T ramps even arrive. Then you have $AAOI and $SIVE appear with scarce, independent CW laser or transceiver capacity... That usually leads forces your other major players like $AMD and hyperscalers to look at them. $MTSI said their capacity only starts to come online H2 2027 and customers are already approaching them with urgency... CW lasers is one of my favorite bottlenecks right now.中文 · 翻译不知道是不是我一个人的感觉，但我现在看到的 CW（连续波）需求失衡太大了（多亏了 $NVDA）。 而且这还没算上规模扩大和其他 1.6T 产能爬坡还没到来呢。然后你看 $AAOI 和 $SIVE 出现了，它们有稀缺的、独立的 CW 激光器或收发器产能…… 这通常会迫使其他主要玩家，比如 $AMD 和超大规模数据中心提供商，去关注它们。$MTSI 说他们的产能要到 2027 年下半年才开始上线，但客户已经很急地来找他们了…… CW 激光器是我现在最喜欢的一个瓶颈。
- English · 原文uhhh... the Unitree IPO is more than 8000 times oversubscribed by retail investors. The demand for pure play humanoid companies is absolutely enormous? I'm personally in the US-based Agility Robotics club (Softbank, $NVDA, $AMZN, Foxconn, etc), at $2.5B premoney via $CCXI. And my take is that Unitree likely opening at $30B+ (from pre-ipo perps), might bring the leading US players some more attention in a week or two. Regardless that's just absurd demand.中文 · 翻译呃……宇树科技这次IPO，散户认购超额了8000多倍。 所以，市场对纯粹做人形机器人的公司，需求简直是海量吗？ 我个人是属于美国那边的 Agility Robotics 俱乐部（有软银、$NVDA、$AMZN、富士康等大佬投资），估值是 $2.5B 投前（通过 $CCXI 参与）。 我个人看法是，宇树科技可能要以 $30B+ 的估值开盘（看它IPO前那些私募份额），这或许能让美国那边领先的几家公司在接下来一两周里获得更多关注。 不管怎么说，这需求简直太离谱了。
- English · 原文Just some near term events: - OCP APAC tomorrow (Ayar, Lightmatter, $AMD, $NVDA) and your CPO players are giving announcements/updates. This should be a catalyst for certain optical players. - Earnings week: with $ASTS, $RKLB reporting today (space), $LITE on Tuesday (photonics), $NBIS in the middle of the week (Neocloud), and more. - Unitree IPO subscriptions opened up today and Listing is expected this month. Should be a potential catalyst for the humanoid + robotics sector if it opens up well eg. $CCXI. Fun week ahead.中文 · 翻译就说几个近期的事件吧： - 明天就是OCP亚太峰会（Ayar, Lightmatter, $AMD, $NVDA），届时你们关注的CPO玩家会发布公告/更新。 这应该会对某些光模块公司形成催化剂。 - 财报周：今天有$ASTS、$RKLB（航天领域）发财报，周二有$LITE（光子学领域），周中有$NBIS（Neocloud），等等。 - Unitree的IPO申购今天开放了，预计本月上市。 如果它上市表现不错，这应该会对人形机器人+机器人板块形成一个潜在的催化剂，比如$CCXI。 接下来一周会很有趣。
- English · 原文I'm extremely bullish on $AAOI, it's one of my top holdings? But there's an unholy amount of things to unpack when AOI said they're not participating in first-gen CPO deployments because there's not enough capacity. 1. Independent CW DFB capacity, even less than markets thought. 2. Qualified CPO laser players, even less than markets thought. 3. I remember Rosenblatt thought $AAOI would be $AMD's first CPO customer, but might not be the case anymore... And since $LITE and $COHR are in camp $NVDA and sold out for next 2 years... Some other CPO players might be more valuable than others think.中文 · 翻译我对 $AAOI 非常看好，这是我持仓里的大头？ 但当 AOI 说他们不会参与第一代 CPO 部署，因为产能不足时，里面有惊人多的东西要细说。 1. 独立的连续波 DFB 产能，比市场预期的还要少。 2. 合格的 CPO 激光供应商，比市场预期的还要少。 3. 我记得 Rosenblatt 认为 $AAOI 会是 $AMD 的第一个 CPO 客户，但现在可能不再是这样了... 而且因为 $LITE 和 $COHR 都在 $NVDA 阵营，未来两年订单都满了... 其他一些 CPO 玩家可能比大家想象的更有价值。
- English · 原文Funniest callout of the year from $VIAV on CPO timelines: "Well, there's been a lot of industry talk [about delays], like, 'Oh, because the yield is going to be slower.' That's all nonsense. CPO and all that thing is moving forward" They stated that Yes there's always issues, but the process is being improved and things are getting better. To follow up on that, the CEO stated: - Viavi has POs for CPO-testing activity - “We’re already getting [CPO revenue] this quarter, and probably in December it will start accelerating.” On a side note, they also stated, 1.6T is "ramping very quickly", should reach parity with 800g next year, and signaled longer tail for 800g demand. TLDR: Seeing CPO test revenues hit for $VIAV to $FORM around now. Nvidia also stated switch-side CPO for scale out and scale across has entered mass production earlier. As for scale up timelines, early production should be H2 2027 for some players, with other CPO players Ayar stating 2028 is the volume inflection period. Basically, all the timelines seem to be the same as before and revenue is starting to hit testing players first.中文 · 翻译$VIAV 谈到 CPO（共同封装光学）的时间表时，今年最搞笑的一段话： “好吧，行业里有很多关于（延迟）的讨论，比如‘哦，因为良率会比较慢。’ 那都是胡说八道。CPO 和所有相关的东西都在向前推进。” 他们表示，是的，总会有问题，但流程正在改进，情况正在好转。 接着，CEO 进一步表示： - Viavi 已经收到了 CPO 测试活动的采购订单 - “我们这个季度就已经收到（CPO 收入）了，可能到 12 月就会开始加速。” 另外，他们还提到，1.6T “正在快速爬坡”，预计明年将与 800G 持平，并预示 800G 的需求会有更长的“长尾”效应。 总结：$VIAV 正在为 $FORM 等公司带来 CPO 测试收入。英伟达之前也说过，用于横向扩展和跨域扩展的交换侧 CPO 已经进入了大规模生产。 至于纵向扩展的时间表，一些厂商的早期生产应该在 2027 年下半年，而其他 CPO 厂商，比如 Ayar 公司，则表示 2028 年是产量拐点。 基本上，所有的时间表似乎都和之前一样，收入正开始首先流向测试设备厂商。
- English · 原文As a earnings recap: $AMZN, $META, $GOOGL, and $MSFT guided a 2026 combined capex forecast to ~$720-$745 Billion. Up from $695-$725B Billion previously. Amazon: $220B Google: $195B-$205B Meta: $130B-$145B Microsoft: $175B We've already seen significant deleveraging and retail/institutional margin liquidations (maybe it continues for more time, who knows). But medium-long term, I'm not quite sure how anyone can be bearish the upstream semis or neoclouds. Given each hyperscaler has flagged either compute shortages, rising cloud demand + pricing power, or increased spending for chips/networking. My "bottleneck" thesis with many of these upstream semi supply chains is that when trillions in capital flows into things from InP substrates or memory (which were both treated as cheap commodities) or even energy. Lot of these current AI names that were treated as useless before in telecom cycles or even toilet sellers. Gets rerated when their inflection period hits and capex flows into their balance sheets. As seen with $NVDA GPUs past few years, memory this year, CPUs/MLCCs next few years, CPO in 2027, Glass Substrates in 2027, 800V in 2027, and so on.中文 · 翻译财报总结一下：$AMZN、$META、$GOOGL 和 $MSFT 预计，到2026年，它们的合计资本开支将达到约7200亿至7450亿美元。 比之前预期的6950亿到7250亿美元有所提高。 亚马逊：2200亿美元 谷歌：1
- English · 原文Just some interesting takeaways from $FORM earnings call on CPO: Since they attributed systems revenue nearly doubled due to: "accelerating growth in co-packaged optics or CPO". Q1 2026: "we now expect 2026 CPO revenues to come in at the high end of the $10 million to $20 million range" Q2 2026 (now): "expect to exceed that range by the end of the third quarter, and to significantly exceed the $20 million level for the year overall." Management stated "We’re seeing some significant acceleration in this [CPO] business". Q: CPO adoption outlook. Are we perhaps expecting the timeline to accelerate a little bit? A: "What I would say is there’s acceleration here in the very short term." But management stated there's not a significant pull-in on timelines. Citing production infrastructure (not just R&D) for: - The growing volumes of CPO chips planned for later this year + test insertion for scale-up and scale-out switches. - Seems like that CPO piece flows through $TSM COUPE [ likely maps to $NVDA CPO products] - "The CPO piece does flow through that 10% customer" So now, test is ramping given management cited growing volumes of CPO chips. "The rapid recent growth of our CPO business is an exciting development, which we believe represents the very early stages of widespread adoption of silicon photonics in the broader semiconductor industry" TLDR: Fundamentally, CPO as a theme is very early, and should start to go brrr soon since it's starting to show up in earnings now (equipment/test players usually appear up first before optical engines / laser volumes ramp). Personally once all the deleveraging stops, I think markets will start to care more.中文 · 翻译整理了 $FORM 财报电话会上关于 CPO 的一些有意思的看点： 因为他们将系统收入几乎翻倍归因于：“共同封装光学器件（CPO）的加速增长”。 2026 年第一季度（预测）：“我们现在预计 2026 年 CPO 收入将达到 1000 万至 2000 万美元区间的上限。” 2026 年第二季度（最新预测）：“预计到第三季度末将超出该范围，全年 CPO 收入将显著超过 2000 万美元。” 管理层表示：“我们看到这项 CPO 业务正在显著加速。” 问：CPO 的采用前景如何？我们是否可以期待时间表能稍微提前一些？ 答：“我想说的是，短期内确实有加速。”但管理层表示，时间表没有显著提前。 提到需要生产基础设施（不仅仅是研发），是为了： - 今年晚些时候计划的 CPO 芯片产量增长 + 用于横向扩展和纵向扩展交换机的测试导入。 - 看起来 CPO 这块业务是通过 $TSM COUPE 走的 [ 很可能对应的是 $NVDA 的 CPO 产品 ] - “CPO 这块业务确实通过那个占我们 10% 收入的客户。” 所以现在，考虑到管理层提到了 CPO 芯片的产量正在增长，测试环节正在加速。 “我们 CPO 业务近期快速增长是一个令人振奋的进展，我们相信这代表了硅光子技术在更广泛半导体行业中普遍采用的早期阶段。” TLDR（太长不看）：从根本上看，CPO 作为一个主题还处于非常早期，但它已经开始在财报中显现（设备/测试厂商通常在光引擎/激光器出货量大增之前就先有表现），所以应该很快就要起飞了。 就我个人而言，一旦所有的去杠杆化停止，我觉得市场就会开始更关注它了。
- English · 原文Just some takeaways from $META | $MSFT earnings calls: Microsoft: - Expected to be FCF positive in 2027, despite the increase in capex (extremely positive for AI buildout that it's funded by operating income) - "Free cash flow was $19.6 billion, reflecting higher capital expenditures" - Quartely capex was $41B, roughly 2/3rds were "short lived assets, primarily CPUs and GPUs" - Expects capex spend will be over $50 billion for next quarter - Capex Guidance at ~$175 billion and 2027 capex roughly the same. Spending plans unchanged and in line. - "Extending the estimated useful life of our data centers from 15 to 25 years" - "We will be among the first cloud providers to deploy next generation rack-scale AI infrastructure based on $AMD Helios and $NVDA Vera Rubin" - "Customer demand continues to exceed available capacity" Meta: - Capex $130-$145 billion (narrowed range), from $125B-$145B. - Meta is receiving offers at a "significant premium" to what they paid for it (compute scarcity, positive for neoclouds like $IREN / $NBIS ) - Expects significant portion of compute (like the 1 GW DC in El Paso) to develop internal models. - Meta has multiple ROI-positive uses for additional compute across its core business (internally, not Meta Compute) - "Finally, we believe that overall industry capacity is going to remain tight for the foreseeable future" - "The industry has under-built historically for the wave of AI adoption, making existing capacity, including our own, extremely valuable" - Susan Li TLDR: - $MSFT and $GOOGL largely sustaining AI capex buildout while remaining FCF positive or through operating incomes. - $META flags available compute materially below demand at least through 2027. And $MSFT also flags compute demand far exceeds supply. - All three hyperscaler capex largely in line with Google hiking capex figures. AI selloff seems extremely overblown now, hyperscalers continuing capex in line (with Microsoft being FCF positive) or even hiked with $GOOGL. Compute scarcity is visible throughout every single hyperscaler ER.中文 · 翻译简单总结一下 $META 和 $MSFT 财报电话会的几点收获： 微软： - 预计到2027年自由现金流（FCF）转正，尽管资本支出（capex）在增加（这对于AI基础设施建设是天大的利好，因为它的资金来源是运营收入） - “自由现金流为19
- English · 原文Nvidia in talks to backstop $250B worth of OpenAI DC financing. - Softbank’s SB energy develops 10 GW Ohio campus - OpenAI signs long term lease - $NVDA guarantees $250B in financing - OpenAI discussing separate agreement to buy up to $350B in Nvidia chips. Not quite sure if this circular financing is healthy long term, but given the massive size of AI infrastructure spending… Feels like capex beneficiaries will be extremely happy.中文 · 翻译Nvidia 正在洽谈，计划为 OpenAI 价值 2500 亿美元的数据中心融资提供担保。 - 软银旗下的 SB energy 在俄亥俄州开发 10 GW 的园区 - OpenAI 签署长期租赁协议 - $NVDA 担保 2500 亿美元融资 - OpenAI 正在讨论另外一份协议，计划购买高达 3500 亿美元的 Nvidia 芯片。 不太确定这种循环融资长期来看是否健康，但考虑到人工智能基础设施的巨大开支…… 感觉资本支出受益者们会非常开心。
- English · 原文Jensen from $NVDA has joined X for the first time to share his letter on open models.中文 · 翻译英伟达（$NVDA）的黄仁勋首次来 X 分享他关于开放模型的信。
- English · 原文$AMD to invest $5B in Anthropic. Anthropic in turn will sign a 2 GW deal to buy AMD’s latest chips, worth tens of billions of dollars. Looks like they’re copying the $NVDA playbook with Anthropic last year. https://t.co/SXBUq65u6P中文 · 翻译$AMD 将向 Anthropic 投资 50 亿美元。 Anthropic 反过来会签一份 2 吉瓦的协议，购买 $AMD 最新的芯片，价值数百亿美元。 看起来他们正在复制 $NVDA 去年和 Anthropic 合作的那一套玩法。https://t.co/SXBUq65u6P
- English · 原文It's slightly more nuanced. Basically, $AMD SF event is tomorrow, Wed/Thurs (July 22, 23). From Jefferies client note, AMD is expected to provide additional disclosures around their likely scale-up CPO roadmap + MI500. If they confirm CPO route, would positive thematically since it's not just $NVDA forcing the transition. As for $SIVE, we'll get a better read if they're directly in AMD supply chains. Since right now, it's possible to map it through $GFS (reportedly involved in AMD MI500 CPO work) or AYAR (which AMD invested in). But no formal confirmations. So if you're turning in: - CPO thematic positive: MI500 explicitly commits to CPO/optical UALink. - AMD gives details using Ayar Labs or GF SCALE for MI500 CPO: That would be extremely positive for $SIVE, given they're the upstream laser supplier / reference laser. Currently, Blayne Curtis from Jefferies expect AMD to work with $ALAB on UAL and $AVGO on ESUN... but juicy details are here: "We are looking for whether AMD commits to a CPO based scale up solution and who supplies the optical engine" We'll get a broader read through on the supply chain come Wednesday/Thursday, too early to tell today.中文 · 翻译这事儿其实更复杂一些。简单来说，$AMD 的旧金山活动就在明天，周三/周四（7 月 22、23 日）。 根据 Jefferies 的客户报告，$AMD 预计会披露更多关于他们可能的大规模 CPO（共同封装光学）路线图 + MI500 的信息。 如果他们确认走 CPO 路线，那从主题上来说是利好，因为它表明不只是 $NVDA 在推动这一转变。 至于 $SIVE，如果他们直接进入 $AMD 的供应链，我们就能了解得更清楚。 因为目前，我们有可能通过 $GFS（据报道参与了 $AMD MI500 CPO 工作）或 $AYAR（$AMD 投资的公司）来推断，但还没有官方确认。 所以如果你正在关注： - CPO 主题积极面：MI500 明确承诺采用 CPO/光学 UALink。 - $AMD 透露 MI500 CPO 将使用 Ayar Labs 或 GF SCALE：考虑到 $SIVE 是上游激光供应商/参考激光，这对于 $SIVE 来说将是极其积极的。 目前，Jefferies 的 Blayne Curtis 预计 $AMD 将与 $ALAB 合作 UAL，并与 $AVGO 合作 ESUN……但关键细节在这里： “我们正在关注 $AMD 是否承诺采用基于 CPO 的扩展解决方案，以及谁将提供光学引擎。” 我们将在周三/周四获得更广泛的供应链信息，今天下结论还为时过早。
- English · 原文If you’re curious why $NBIS is up after hours. $NVDA disclosed it owns 9.3% beneficial ownership of Nebius via SEC filings. Not exactly too new since it dates back to Nvidia’s existing share position of 1.19M shares + $2B prefunded warrant of ~21M shares. But great for sentiment having Nvidia be a large shareholder of the Neocloud leader.中文 · 翻译如果你好奇为什么 $NBIS 盘后涨了。 $NVDA 披露，它通过 SEC 备案文件拥有 Nebius 9.3% 的实益所有权。 其实这不算什么太新的消息，因为这可以追溯到英伟达之前持有的 119 万股股份，以及一份 20 亿美元的预付认股权证（对应约 2100 万股）。 但对于市场情绪来说，英伟达能成为 Neocloud 领导者的大股东，这是个好消息。
- English · 原文Agreed! It’s nice to remember your thesis during a market crash. From my own personal thesis, if $AAOI hits $1.4B quarterly revenue start of Q3 2027. Which is annualized $5.6B off a $8B MC. Is it “over” for the company if that revenue ramp hasn’t even shown up in the quarterly earnings… when it’s 2026? Same applies to my CPO sector exposure like $SIVE, an architecture shift led by $NVDA. If scale out volume ramps from H2 2026 into 2027, and scale up heavily volume ramps into 2028. Is it “over” that $0 -> $91B TAM expansion (per GS) hasn’t even hit yet? With robotics like Agility, is it over in 2026 if the listing hasn’t even happened yet and humanoids haven’t ramped? I personally think current market conditions are a reflection of liquidity and leverage, not individual fundamentals. It’s brutal for everyone to see KOSPI, TW, Nikkei, and AI, space, robotics sector stocks crash recently. Especially when there’s a lot of irrational behavior stemming from those leverage. In the end, we can’t tell you what stocks to buy, what timeframe you should sell, how to size your positions, or what you should do. Only share personal thoughts or research and track if they get validated over time. So it’s extremely important to build your own thesis, since everyone has unique risk tolerance or investing timeframes. And that usually leads to having higher conviction during crashes.中文 · 翻译同意！在市场崩盘时还能坚守自己的投资理论，感觉真好。 从我个人投资理论来看，如果 $AAOI 在 2027 年第三季度初达到 14 亿美元的季度营收。这相当于 80 亿美元市值下，年化营收达到 56 亿美元。 如果到 2026 年，这种营收的快速增长还没体现在季度财报中……那这家公司就算“完了”吗？ 这同样适用于我在 CPO（协同封装光学）板块的持仓，比如 $SIVE，这是由 $NVDA 引领的架构转变。 如果从 2026 年下半年到 2027 年，横向扩展（scale out）
- English · 原文$NVDA CEO, Jensen Huang in Tokyo today said: "The reports are not true. Vera Rubin is already in production. Giant amounts of production incoming" As a flat denial to the accusation that their AI servers would be delayed. Jensen appears to be using the Umbrella term for Rubin Ultra/Kyber, then pointed to current gen ramp. Regardless, lot of damage has been done already to Nvidia supply chains stock prices, after the multiple delay reports. Even if Nvidia refuted them by clarifying Kyber change was an architecture optimization, not impacting timelines. It's a very dangerous and growing trend for folks to post overreaching broader claims from technical nuances to grow viewcount... Especially if they get Jensen has to respond. In the end, I'm trusting $NVDA on timelines since they probably have the greatest visibility into their own supply chain.中文 · 翻译$NVDA CEO 黄仁勋今天在东京表示： “那些报道不属实。Vera Rubin 已经投产了。正在大量出货。” 这是对他们 AI 服务器会延期的指控的断然否认。 黄仁勋似乎是在用一个统称来指代 Rubin Ultra/Kyber，然后又提到了现有产品的产能提升。 不过，在多份延期报告之后，对英伟达供应链和股价的损害已经造成了。即便英伟达澄清说 Kyber 的变化是架构优化，不影响时间表，也一样。 现在，有这么一种非常危险且日益增长的趋势：有些人从技术细节中引申出过于宽泛的说法，就为了增加阅读量……尤其是如果这能让黄仁勋都不得不出来回应的话。 归根结底，我还是相信 $NVDA 在时间表上的说法，毕竟他们对自己的供应链肯定是最清楚的。
- English · 原文It's nuanced, $AEHR implications was more toward broader silicon photonics related ramp. Which could include 800g/1.6T pluggables, or CPO. If they directly named CPO, Msscorps would be a very clean readthrough. It just got hit really hard recently from that delay claim (even after $NVDA refuted it), since it's it's upside seems tied to CPO yields in specific. So it's positive, just uncertain.中文 · 翻译这事儿有细微之处，$AEHR 的影响更偏向于推动更广泛的硅光子学相关增长。 这可能包括 800G/1.6T 的可插拔光模块，或者 CPO（共封装光学）。 如果他们直接点名 CPO，那 Msscorps 就会是一个非常明确的受益股。它最近因为那个延迟的传闻被砸得很惨（即使 $NVDA 已经辟谣了），因为它股价的上涨似乎特别依赖于 CPO 的具体良率。 所以这是个利好，只是不确定性比较高。
- English · 原文So Jensen himself went out to do damage control with $NVDA delays reports apparently: 1. Jensen Huang said no delays: - 800V and rack-to-rack optical interconnects is progressing fully in line with the original plan with no material delay. - Debunked the rumor that the next-generation flagship, Rubin Ultra, would be delayed until 2028 and said it would be on track to ship next year - Rack system optimizations, like replacing Kyber with a new design are architectural optimizations, not impacting delivery timelines. 2. On ASIC competition: no loss of market share. - Talked about how a leading frontier lab that previously relied heavily on internal ASICs has shifted nearly 50% of its compute load to Nvidia 3. On over-reliance on hyperscalers. - AI frontier labs currently contributing to ~20% of demand. - Traditional cloud providers, Nvidia provides Vera CPUs and high-speed networking - Sovereign AI中文 · 翻译看来黄仁勋亲自出来，就$NVDA延期报道进行危机公关了： 1. 黄仁勋说没有延期： - 800V和机架间光互连的进展完全符合原计划，没有实质性延误。 - 驳斥了下一代旗舰Rubin Ultra将推迟到2028年的谣言，并表示会按计划于明年出货。 - 机架系统的优化，比如用新设计取代Kyber，这些都是架构优化，不会影响交付时间。 2. 关于ASIC竞争：没有市场份额损失。 - 提到一个之前严重依赖内部ASIC的顶尖前沿实验室，已将其近50%的计算负载转移到英伟达。 3. 关于过度依赖超大规模客户的问题。 - AI前沿实验室目前贡献了约20%的需求。 - 传统云服务商方面，英伟达提供Vera CPU和高速网络。 - 主权AI
- English · 原文Apparently there’s a new popular trend to bearpost trillion dollar companies. Like $NVDA or $TSM. By finding isolating technical details, then blowing it out of proportions by claiming their whole program is delayed. -> if the company doesn’t respond, stock prices/sentiment are damaged. -> if the company issues a response, they can’t go into detail since supply chains are confidential… but people still doubt them -> Poster gain more impressions, then does it again with more views the next time. Don’t quite think this is very healthy, with more folks copying that playbook. Especially around incentive structures.中文 · 翻译貌似现在流行一种新趋势，就是专门去唱空那些万亿美元市值的公司。 比如 $NVDA 或 $TSM。 具体做法就是抓住一些孤立的技术细节，然后夸大其词，声称他们的整个项目都被延误了。 → 如果公司不回应，股价或市场情绪就会受损。 → 如果公司做出回应，又不能详细说明，因为供应链是保密的……但人们还是会怀疑他们。 → 发帖人获得更多曝光，下次会获得更多关注，然后又这么做。 我不觉得这很健康，因为越来越多的人开始效仿这种套路。尤其是在激励机制这块。
- English · 原文Apparently, Korean Beauty was where all the alpha is. APR (278470.KS) is up 529% over the past 2Y, wildly outperforming both $NVDA and $INTC. Every female I know talks about their "Medicube" stick, then buying products... where you put a whole salmon's family tree on your face for better skin. The world's female population creates a bigger structural demand than hyperscalers for memory or lasers? Should have known...中文 · 翻译看来，韩流美妆才是真正的超额收益所在啊。 APR (278470.KS) 过去两年涨了 529%，表现远超 $NVDA 和 $INTC。 我认识的每个女生都在谈论她们的 "Medicube" 棒，然后买那些产品……就是那种，把一整条三文鱼的祖宗十八代都抹到脸上，就为了皮肤变好。 全球女性人口产生的结构性需求，居然比超大规模云计算公司对内存或激光的需求还要大？早该知道的嘛…
- English · 原文He wouldn't have build a $1T company without pulling off all those moves. - Gets funding from Elon -> Pisses off $TSLA going independent - Gets funding from $MSFT -> Pisses of Microsoft by diluting them with other hyperscaler funding - Gets funding from $NVDA -> Builds their competing chip and funds Cerebras competitors for inference - Works with $AAPL on Apple Intelligence? -> Create their competing devices. - Potential lack of funding? -> Sign a lot of LTAs so the entire industry lives or dies off OpenAI succeeding. Whatever he's doing is working lol.中文 · 翻译他要是没做成所有这些操作，根本不可能打造出一个万亿美元公司。 - 从马斯克那里获得资金 → 激怒 $TSLA，走独立路线 - 从 $MSFT 那里获得资金 → 通过引入其他超大规模厂商的资金稀释微软的股份，惹恼了微软 - 从 $NVDA 那里获得资金 → 打造他们自己的竞争芯片，并为 Cerebras 的推理竞争对手提供资金 - 和 $AAPL 合作开发 Apple Intelligence？ → 然后制造他们自己的竞争设备。 - 潜在的资金短缺？ → 签署大量长期协议，让整个行业都与 OpenAI 的成败休戚相关。 不管他怎么做，反正都成功了，哈哈。
- English · 原文I personally think so. It feels like markets started selling off after that misleading Bloomberg $META compute article, and a couple of SemiAnalysis bear posts on $NVDA supply chains. Since algos + headline interpretation of “Meta excess compute” was cutting capex + dropping out of AI race. Combine that with excess leverage and here we are. Probably just takes an earnings call and few Trump tweets to get back to normal… But after some names already dropped 40-60% feels like it should bottom soon. Reversals the other way are pretty violent too.中文 · 翻译我个人觉得是这样。感觉市场开始抛售，是在那篇误导性的彭博社 $META 算力文章出来之后，还有SemiAnalysis发的几篇看空 $NVDA 供应链的文章。 因为算法交易加上对“Meta算力过剩”这种标题的解读，就变成了Meta要削减资本支出，退出AI竞赛。 再加上过度的杠杆，就到了现在这个地步。也许只需要一次财报电话会议和几条特朗普的推文，就能恢复正常了…… 但一些股票已经跌了40-60%了，感觉应该很快就要见底了。 反过来说，反弹起来也通常很猛。
- English · 原文Be Samsung at $1.24T: > market: we don’t think you can keep hiking memory prices > proceeds to hikes dram by 20% > releases earnings > most profitable company in the world beating $NVDA and $AAPL > operating profits growing 1803% Y/Y market: sells off Samsung -7%中文 · 翻译想象一下，三星市值1.24万亿美元的时候： > 市场：我们觉得你不能再涨存储器价格了。 > 结果它转头就把DRAM价格涨了20%。 > 然后发布财报。 > 成了全球最赚钱的公司，业绩超过了$NVDA和$AAPL。 > 营业利润同比增长1803%。 市场：然后把三星股价砸了7%。
- English · 原文One of the dumbest thematic selloffs I’ve seen to to date off: - $META compute news that’s not new - CPO delay report 1, that got refuted by $NVDA - CPO delay report 2, that got refuted by $NVDA High confidence, institutions will end up going long on the same names they’re bearposting after retail capitulates.中文 · 翻译这是我迄今为止见过的最愚蠢的主题性抛售之一，起因是： - $META 的计算相关新闻，但那根本不是什么新消息 - CPO 延迟报告 1，然后被 $NVDA 辟谣了 - CPO 延迟报告 2，结果又被 $NVDA 辟谣了 我很有信心，散户彻底投降后，机构最终会做多他们之前一直在发帖唱空的那些股票。
- English · 原文It’s interesting to witness psychology around valuation anchoring + scarcity. Retail, for example, are buying Figure, last valued at ~$39B through CEFs… At $158B, since it’s private, round 4x valuations. Then Agility Robotics $CCXI, which has broader commercialization. Is at ~$4.3B pre-money and publicly available before the name change. At closer pricing that $AMZN, $NVDA, SoftBank, and Foxconn vaulted it at. Yet some end up paying 4-5x prices of already hiked private investor rounds for exposure… I’m convinced if Agility raised a very small private round at $39B as well from $2.5B (which it definitely could), to set a valuation anchor. Investors would be foaming in the mouth for a private allocation, just due to psychology rather than underlying fundamentals. We’ve also seen this anchoring with $SPCX recently at $1.75T. Peak market inefficiency? Or lack of knowledge from retail?中文 · 翻译看到估值锚定和稀缺性引发的心理现象，真有意思。 比如，散户正通过封闭式基金（CEFs）购买 Figure，这家公司上次估值大概是 390 亿美元左右… 在 1580 亿美元的估值上，因为它是私有公司，差不多是（之前）4 倍的估值。 然后是 Agility Robotics $CCXI，这家公司的商业化前景更广。 它的投前估值大概是 43 亿美元，而且在公司改名前是公开可得的。当时的定价，更接近于 $AMZN、$NVDA、软银和富士康给它抬的那个价格。 然而，有些人最终为了获得敞口，支付了比已经涨过价的私募轮高出 4-5 倍的价格… 我确信，如果 Agility 也以 25 亿美元到 390 亿美元的估值进行一轮小规模私募（它绝对能做到），以此设定一个估
- English · 原文Kelper upgrades $XFAB to "Buy" citing: - Melexis demand strengthening “week-after-week, day-after-day” (their lead customer). - AI giving XFAB a structural growth leg with SiC, GaN, and photonics offsetting auto cyclicality. - Scarce Western speciality foundries like $SILEX commanding higher multiples. "X-Fab remains the cheapest specialty foundry peer on every multiple, trading at 4.8x 2027 EV/EBITDA versus a peer median of 14.5x" I treat this as good signal for the core fundamentals improving like auto players. But in terms of unpriced long term growth vectors, I'm personally looking at XFAB from a different angle in terms of photonics growth with photonixfab. Given $NVDA as a potential key driver. But it's hard for traditional firms to price these types of active developments into valuation models until there's volume contracts.中文 · 翻译Kelper 将 $XFAB 评级上调至“买入”，理由是： - Melexis 的需求“周复一周，日复一日地”不断走强（Melexis 是他们的大客户）。 - AI 通过碳化硅 (SiC)、氮化镓 (GaN) 和光子学，正在赋予 XFAB 结构性增长动力，
- English · 原文Let’s just say I don’t think markets view them as objective anymore. There’s been tendencies to post overreaching claims over a nuanced technical details. Like how $MU had no HBM4 share with $NVDA few months back. I’m personally just listening to $NVDA direct statements and roadmap.中文 · 翻译这么说吧，我觉得市场已经不把他们看作是客观的了。 他们一直倾向于对一些细致的技术细节发表过于夸大的说法。比如几个月前，关于 $MU 在 $NVDA 的 HBM4 方面没有份额的说法。 我个人现在就只听 $NVDA 的官方声明和路线图了。
- English · 原文@softmaxedx I blame Semianalysis for crashing the entire photonics market for misleading CPO delay reporting that $NVDA refuted. Alongside global journal $META recent misleading reporting… But I think they realized everything was crashing along with it, so they needed a macro bull post.中文 · 翻译@softmaxedx 我要怪 Semianalysis，他们搞崩了整个光子学市场，就因为那份被 $NVDA 反驳的、误导性的 CPO 延迟报告。 再加上 $META 最近那些全球性的误导性报道… 但我觉得他们意识到所有东西都跟着在崩盘，所以他们需要发一个宏观看多的帖子。
- English · 原文What's this PTSD about SPACs? I only care about the company fundamentals. $NVDA, Softbank, $AMZN and others don't just back some random crapco. And Foxconn is funding the PIPE. It's just for faster listing, especially giving US investors alternatives to the extremely anticipated up Unitree IPO coming out around the same timeframe.中文 · 翻译对 SPAC 怎么会有这种 PTSD？我只关心公司的基本面。 $NVDA、软银、$AMZN 等公司可不是随便投些垃圾公司的。而且富士康正在为 PIPE 提供资金。 这只是为了更快上市，特别是为美国投资者提供一个替代方案，因为备受期待的 Unitree IPO 也在同期上市。
- English · 原文Agility Robotics is literally the #1 humanoid player in the US in terms of commercialization today imo. Personal Rankings: 1. Agility Robotics via $CCXI 2. $TSLA Optimus ($1T+ company) 3. Figure at $39B from 2025, likely higher now. With Tesla likely taking the lead a year from now. So at $16, you have a ~$4B MC pre-money humanoid player. Which is already inside $AMZN and $MELI (ideal use cases to start out, before step-functioning to general-use), with $AMZN, $NVDA, Softbank, Foxconn, and others as investors. NFA, but at those valuations... there's direct exposure for the first ever humanoid player today, that's likely further ahead than Figure for commercialization. No hard feelings about Robostrategy, but it's over 4x+ NAV: Which would value Figure at $156B... While there's the #1 commercialized humanoid player with Agility at $4B today. So my personal preference is $CCXI.中文 · 翻译在我的看法，Agility Robotics 可以说是目前美国在商业化方面排第一的人形机器人玩家。 个人排名： 1. Agility Robotics （通过 $CCXI） 2. $TSLA Optimus （万亿美元市值以上公司） 3. Figure 2025年估值 390 亿美元，现在可能更高了。 特斯拉很可能
- English · 原文Are you stupid? 1. $JBL mass producing 1.6T LRO is revenue. 2 O-Net mass producing ELS is revenue. 3. $GFS using $SIVE as reference laser for the hyperscalers is revenue. 4. Ayar joining $NVDA NVLink for CPO and using $SIVE for mass production is revenue. 5. Other pluggable players, potentially Eoptolink/Innolight mass producing pluggables with $SIVE as lasers is revenue. 6. $POET entering volume production is revenue.中文 · 翻译你是不是傻？ 1. $JBL 量产 1.6T LRO 就是营收。 2. O-Net 量产 ELS 就是营收。 3. $GFS 给超大规模数据中心用 $SIVE 作为参考激光器，就是营收。 4. Ayar 加入 $NVDA 的 NVLink 做 CPO，并用 $SIVE 进行量产，就是营收。 5. 其他可插拔模块厂商，可能是 Eoptolink/Innolight，量产以 $SIVE 作为激光器的可插拔模块，就是营收。 6. $POET 进入量产就是营收。
- English · 原文Photonics is backed by actual revenue numbers and it's an architectural shift championed by $NVDA. Quantum barely has any revenue. $LITE is completely sold out for the next 2 years (per $POET AGM) likely starting into 2029. Lumentum is so strained that they buy CW lasers off competitors (earnings transcript) $COHR is bottlenecked, so they buy EML off Lumentum. Then, $AAOI is coming in with Made-in-America independent CW capacity, are projecting $1.4B/quarterly revenue ending H1 2027 of a stupid $9.3B MC today. So all the CW capacity from independent players who have it now like $AAOI or $SIVE are likely to become scarce resources. Many other hyperscalers have already started LTA discussions (per Trendforce). And players like $AMD are currently talking with players such as $AAOI (Rosenblatt channel checks). Thematically, next 2 years is 9x TAM to US$154B per GS reports, especially with 16x/45x dollar content increase in scale out/scale up. Then there's the overall thematic AI drop from $META, which is widely misunderstood because people conflate what "excess capacity" means. And as UBS mentioned, Meta planning a cloud offering is NOT NEW NEWS. Bloomberg just has a tendency to publish information that causes doom drops across the semi sector like Nvidia export controls a few months back. But I'm familiar with what I'm holding so I'm confident in these numbers playing out. Especially when all the major players are sold out, the fundamentals catch up eventually.中文 · 翻译光子技术是有实实在在的营收数据支撑的，而且这是 $NVDA 大力推动的架构转型。 量子技术则几乎没什么营收。 根据 $POET 股东大会的消息，$LITE 未来两年（可能从2029年开始）的产品已经完全售罄。Lumentum 的产能非常紧张，甚至要从竞争对手那里购买连续波激光器（财报电话会议记录里提到了）。 $COHR 则卡在瓶颈上，所以他们向 Lumentum 购买 EML。 接着，$AAOI 正带着“美国制造”的独立连续波产能入局，他们预计到2027年上半年，季度营收将达到14亿美元，要知道它今天市值才93亿美元，简直太低了。 所以，现在像 $AAOI 或 $SIVE 这样拥有独立连续波产能的厂商，他们的产能很可能会变成稀缺资源。 很多其他超大规模数据中心厂商已经开始谈长期供货协议（来自 Trendforce 的消息）。而像 $AMD 这样的公司，目前正在和 $AAOI 等厂商洽谈（Rosenblatt 的渠道调研显示）。 从主题上看，根据 GS（高盛）的报告，未来两年市场总规模（TAM）将增长9倍，达到1540亿美元，特别是在横向扩展/纵向扩展方面，每美元内容投入将分别增加16倍/45倍。 然后还有来自 $META 的关于 AI 的整体主题性下跌，这被很多人误解了，因为大家混淆了“过剩产能”的真正含义。而且正如 UBS（瑞银）提到的，Meta 计划提供云服务并不是什么新鲜事。 彭博社就是有这种倾向，喜欢发布导致半导体行业全面暴跌的消息，就像几个月前关于英伟达出口管制的那次。 但我很清楚自己持有的是什么，所以我对这些数据会实现很有信心。 尤其是在所有主要玩家都产能售罄的情况下，基本面终究会追上来的。
- English · 原文Just a random thought, leading US humanoid players strangely me of the current LLM dynamics: Agility ( $CCXI ) kinda feels like Anthropic for robotics. With $AMZN and $NVDA heavily backing it, ingrained with $GOOGL Deepmind (like TPUs). And it starts off with enterprise commercialization. Optimus is kinda off doing its own thing like xAI. Supported by Tesla/Elon and his visionary roadmap as usual. Figure is like ChatGPT With Microsoft/OpenAI investing (kinda like Microsoft), then ended up kinda competing them by building their own VlA and setting them behind. But ends up top 2 leaders anyway. Boston Dynamics is Gemini Kinda started the entire humanoids thing like transformers with backflip videos. With R&D supported by $GOOGL but somehow let everyone else leapfrog them in commercialization. And then there’s $NVDA just chilling, silently powering the entire humanoids ecosystem.中文 · 翻译随感，美国领先的人形机器人公司，总让我想起现在大语言模型（LLM）的格局，还挺奇怪的： Agility ($CCXI) 有点像机器人界的 Anthropic。 有 $AMZN 和 $NVDA 大力支持，还融入了 $GOOGL Deepmind 的技术（就像 TPUs）。而且它一上来就搞企业级商业化。 Optimus 有点像 xAI 那样，自顾自地在做自己的东西。 一如既往地有特斯拉/马斯克和他的前瞻性路线图做后盾。 Figure 就像 ChatGPT。 有微软/OpenAI 投资（有点像微软），结果却有点和他们竞争起来，自己搞了 VlA，把他们甩在了后面。 不过反正也成了前两
- English · 原文@SidkMena Dunno, probably just sector selloff. Shunsin recebtly confirmed a direct $TSM COUPE partnership. Then proceeded to crash 15-20%. Don’t think there’s any material developments outside SemiAnalysis bearposting the Nvidia ecosystem over and over https://t.co/jxKIKup6F3中文 · 翻译@SidkMena 不知道，可能只是行业普遍抛售。Shunsin 最近确认直接和 $TSM 有 COUPE 合作。 然后股价就跌了 15-20%。 我觉得除了 SemiAnalysis 反复唱空英伟达生态系统外，应该没啥实质性进展了 https://t.co/jxKIKup6F3
- English · 原文Yep! Agility Robotics is currently my favorite humanoid/robotics position. They're set to be listed on NASDAQ via $CCXI as early as September (per Digitimes). Just for informational purposes: Their V4 humanoid robot is already operating in sites like Amazon. And have V5 slated for mass production next year (they have a 10,000+ /year capacity). Investors include Foxconn, $NVDA, $AMZN, Softbank, and now Serenity. I've been personally waiting for humanoid exposure for awhile (Agility is set to be the first US pure-play listed one) to the point I was actually planning on investing in Unitree's IPO. But I'm glad now I personally have a compelling alternative now since I prefer to invest capital to build up Made in America supply chains (75% of their components are US-sourced from investor desks). There are risks assigned to SPACs such as listing delays, or cancellation. But excited to see what happens next with developments. I do hope this encourages other frontier companies to go public early on.中文 · 翻译好的！Agility Robotics 目前是我最看好的人形机器人/机器人领域的持仓。 他们最早可能在九月通过 $CCXI 在纳斯达克上市（据 Digitimes 报道）。 只是提供信息：他们的人形机器人 V4 已经在亚马逊等地方投入运营。而 V5 计划明年量产（他们有每年超过 10,000 台的产能）。 投资者包括富士康、$NVDA、$AMZN、软银，现在又多了 Serenity。 我个人等人形机器人概念股很久了（Agility 将会是第一家在美国上市的纯人形机器人公司），甚至一度计划投资宇树科技的 IPO。 但我很高兴现在我个人有了一个很有吸引力的替代选择，因为我更喜欢把资金投向建立“美国制造”的供应链（他们 75% 的零部件都来自投资者的美国采购）。 SPAC 会有上市延迟或取消的风险。但我很期待看到后续的发展。 我确实希望这能鼓励其他前沿公司尽早上市。
- English · 原文$GM cuts 1,000 workers and replaces them with 50 robots (Jun 22nd report) Reports also show GM is working on a deal with $NVDA on factory robotics. This is industry validation for robotics, since it likely demonstrates increased opex margins + efficiency from employee automation. Especially from companies like $AMZN to General Motors with assembly/warehouse lines that have high headcount. I think the technology is already here and use cases are already proven. Robotics/Humanoids are just pre-scale right now. The wording from large companies is “assistive robotics”… But in all honesty, it’s here to automate away the human workforce and increase profitability. I don’t think companies want to be paying insurance, salaries, etc. when robots can do a better job, and 24/7, so very sad to see potential job losses in the future from robotics automation. Whether people like it or not; seems like the inflection point for robotics is here.中文 · 翻译$GM 裁员 1,000 人，用 50 台机器人替代（6 月 22 日报道） 还有报道显示，$GM 正在与 $NVDA 洽谈工厂机器人方面的合作。 这对机器人行业来说是实打实的背书，因为这很可能说明，员工自动化提升了运营利润率 + 效率。 尤其是像 $AMZN 到 General Motors 这种装配线/仓库线人头密集的大公司。 我觉得技术已经就位，应用场景也已经得到验证。机器人/人形机器人现在只是还没大规模铺开。 大公司的措辞是"辅助型机器人"……但说实话，它就是来替代人力、提高利润的。 我觉得公司们也不想再掏保险、工资这些成本了，既然机器人能干得更好，还能 24 小时不间断工作。看到未来机器人自动化可能带来的失业，真的很遗憾。 不管人们喜不喜欢；机器人的转折点似乎已经到了。
- English · 原文Not really, I personally still think it's undervalued. It's about MC relative to forward revenue potential. In just that tiny timeframe for $SIVE: - $JBL announced 1.6T LRO with Sivers - $GFS reference laser announcement with $SIVE - $AMD went to $GFS for CPO - $SIVE announced CW with TFLN - Ayar (using $SIVE) raised a bunch for volume ramp and joined $NVDA NVLink ecosystem - AlChip private placement with Amazon, which is Ayar's primary ASIC partner (hinting at Trainium wins) - O-Net new ELS partnership with $SIVE for mass production - $POET new purchase order with a customer that stated their initial clients is a "top-3 hyperscaper" - New CHIPS Act nondilutive funding - "LIDAR customer will ramp production starting Q4 2026," which is likely $AEVA - New information discovery likely linking $SIVE customer to $MRVL Celestial (not through Poet), Lightelligence, and Lightelligence. - New information discovery about Ayar removing Lumentum / Macom from website and likely making Sivers primary source - More pluggable players on top of $JBL. - $NVDA bottlenecking CW laser supply and $AMD + other CSPs going around signing LTAs - NASDAQ listing planned And more.中文 · 翻译倒也不是，我个人还是觉得它被低估了。关键是看市值和未来营收潜力的对比。 就在这么短的时间内，关于 $SIVE 就有： - $JBL 宣布与 Sivers 合作，推出 1.6T LRO。 - $GFS 发布了与 $SIVE 相关的参考激光器公告。 - $AMD 转而向 $GFS 寻求 CPO（封装光学）。 - $SIVE 宣布与 TFLN 合作推出 CW（连续波）。 - Ayar（使用 $SIVE 的技术）获得了大量融资以扩大生产，并加入了 $NVDA 的 NVLink 生态系统。
- English · 原文oh i think u misunderstand... I'm actually more bullish than ever as prices go down. $SIVE at ~$1.9B MC you have: - $GFS reference laser - laser for Ayar and others in $NVDA NVLink ecosystem for CPO scale up - $POET and others for CPO scale out - $JBL and others for 1.6T+ pluggable optical transcivers - O-Net for ELS mass production So you have all these hyperscaler suppliers trying to create as much as possible with Sivers as the laser bottleneck... And it's just a matter of how much Win Semi + others partners can make, with $SIVE receiving ~60% gross margins + optionality to TAM expand downward. The 15% recent share expansion that Swedish media have been bearposting is for NASDAQ listing float + M&A (and it's authorization). You just have potential short sellers running illegal bot farms, Swedish media bearposting a company Swedish exchange, on top of a general macro drop. The ~$140M convertible note is real, but it's pennies to US institutions. We'll likely see more institutional ownership when data comes out. But this is also why it's good for $SIVE to prioritize NASDAQ listing so they don't need to deal with this noise and local media. With $AAOI at $10B, you have: - A end-to-end US laser/design/assembly player projecting $471m month revenue H1 entering H2 2027. - With other hyperscalers like $AMD apparently discussing LTAs. And they probably are sitting on a ton of cash after running ~$1.4B in ATMs (400M + 400m+ 600m). Kinda all that needs to be said with those insane revenue projections as long as management doesn't BS. This just reminds me of when $NBIS crashed to $70 last year even while projecting $7-9B ARR, and as timelines got closer it recovered to $250+. Feels like dejavu. I'm just waiting for both volume ramps to hit. Photonics are generally more volatile than the rest.中文 · 翻译哦，我觉得你误会了……随着价格下跌，我反而比以往任何时候都更看好。 $SIVE 市值大约 19 亿美元，它有： - $GFS 的参考激光器 - 为 Ayar 及其他 $NVDA NVLink 生态系统中的 CPO 扩容提供激光器 - 为 $POET 及其他公司提供 CPO 规模扩展 - 为 $JBL 及其他公司提供 1.6T+ 可插拔光收发器 - O-Net 提供 ELS 大规模生产 所以，所有这些超大规模数据中心供应商都想尽可能多地生产，而 Sivers 是激光器的瓶颈…… 这只是 Win Semi 和其他合作伙伴能生产多少的问题，$SIVE 能获得大约 60% 的毛利率，还有向下拓展潜在市场（TAM）的选择权。 瑞典媒体一直在唱衰的最近 15% 股票增发，是为了纳斯达克上市流通股和并购（以及授权）。 你只是看到潜在的做空者在运营非法机器人农场，瑞典媒体对一家瑞典上市公司唱衰，再加上整体宏观下跌。 那笔大约 1.4 亿美元的可转债确实存在，但对美国机构来说只是九牛一毛。等数据出来后，我们可能会看到更多的机构持仓。 但这也就是为什么 $SIVE 优先考虑在纳斯达克上市是件好事，这样他们就不用处理这些噪音和当地媒体了。 $AAOI 市值 100 亿美元，它有： - 一家端到端（美国）的激光/设计/组装公司，预计在 2027 年下半年进入上半年，营收将达到 4.71 亿美元。 - 据说还有像 $AMD 这样的其他超大规模数据中心公司正在讨论长期协议（LTAs）。 而且他们在进行了大约 14 亿美元的 ATM（4 亿 + 4 亿 + 6 亿）后，可能手头现金充裕。 只要管理层不瞎扯，有了那些惊人的营收预测，差不多就够了。 这让我想起了去年 $NBIS 跌到 70 美元的时候，即使当时它预计年度经常性收入（ARR）有 70-90 亿美元，随着时间线临近，它又恢复到 250 多美元。感觉像是似曾相识。 我只是在等待两者的销量都能放量。光子学通常比其他行业波动更大。
- English · 原文$SIVE + Aeva likely coded from the new PR today and SIVE + Apple. For Sivers lasers powering robotics / physical AI. -> Sivers lasers powering Boston Dynamics Atlas via. LG Innotek for robotics -> Nvidia Hyperion ecosystem w/ Sivers lasers for physical AI. Are possible paths with Aeva. Apple likely using Sivers also reiterated from the newer webpage: “Optical sensing modules for wearables” from biometric and health monitoring. They just can’t namedrop partners due to confidentiality, but consumer sensing is probably the biggest tell. Physical AI and mass consumer production with Apple would be very material to revenue.中文 · 翻译$SIVE + Aeva 大概率是从今天的新 PR 里推断出来的，以及 $SIVE + Apple。 关于 Sivers 激光器为机器人 / 具身 AI 提供动力。 → Sivers 激光器通过 LG Innotek 为波士顿动力 Atlas 提供动力，用于机器人领域 → Nvidia Hyperion 生态系统中搭载 Sivers 激光器，用于具身 AI。 这些都是与 Aeva 可能展开的路径。 Apple 也很有可能使用 Sivers，这在较新的网页上得到了重申： “用于可穿戴设备的光学传感模块”——来自生物识别与健康监测。 他们只是出于保密原因不能直接点名合作伙伴，但消费级传感大概是最明显的信号。 具身 AI 加上与 Apple 的大规模消费级生产，对营收来说会是非常实质性的推动。
- English · 原文$XFAB upgraded to outperform with a PT of €12.8 by Bernstein, up from €5. Okay I forgive you Bernstein for Kioxia and Intel. I still think that’s just reactive PTs off automotive recovery, SiC/power semi ramp up. And undershooting potential a lot… If markets price in the possibility of xfab moving to HVM. On their silicon photonics foundry w/ photonixfab / Nvidia + LIGENTEC for TFLN on SOI. Regardless I’m bullish too on XFAB and curious where it heads.中文 · 翻译$XFAB 被 Bernstein 上调至跑赢大盘，目标价从 €5 升至 €12.8。 好吧，我原谅你 Bernstein，在铠侠和英特尔这事上。 我还是觉得这不过是跟着汽车复苏、SiC/功率半导体爬坡做的被动目标价调整。 而且可能远远低估了潜力……如果市场开始计价 XFAB 进入 HVM（高量产）的可能性的话。 毕竟他们在硅光电子代工这块，跟 photonixfab / Nvidia + LIGENTEC 合作搞 TFLN-on-SOI。 不管怎样，我也看好 XFAB，好奇它会走向何方。
- English · 原文@Joey_TheFarmer My thesis hasn't changed with FOCI, they'll be part of the bottleneck with FAU + passive components in $NVDA $TSM ecosystem as that COUPE scales up.中文 · 翻译@Joey_TheFarmer 我对 FOCI 的看法没有改变，随着 COUPE 规模扩大，他们将成为 $NVDA $TSM 生态系统中 FAU + 被动元件瓶颈的一部分。
- English · 原文@eldan_0123 @CreatorNim Don’t think so. $JBL + $SIVE was known from OFC conference in person. But it took an official PR for markets to know. If there any release over the more higher confidence mapping to $NVDA cpo ecosystem, I’d expect it to rerate a lot more中文 · 翻译@eldan_0123 @CreatorNim 不觉得。$JBL + $SIVE 是从 OFC 现场会议里得知的。 但市场要等官方 PR 出来才知道。 如果后续有发布跟 $NVDA CPO 生态更确定的相关消息，我预计它会重新定价不少。
- English · 原文We're in a massive EML bottleneck right now and CW lasers are getting bottlenecked now too. To my knowledge, $COHR is buying EMLs off $LITE because they can't make enough. $LITE is majority allocated to EML, so they can't make enough CW lasers. So they're buying CW lasers off competitors from their ER transcript, (probably Sumitomo, Furukawa and others), which likely feeds into $NVDA contracts. Then $AMD and your hyperscalers need capacity too but we're already in a shortage. Any capacity that comes online would likely be bought, since looks like we're in a shortage for next few years (lumentum already sold out into 2028). Not quite the same as your sk hynix/samsung/micron memory dynamic but there's still a massive moat with EML with probably only single digit amounts of players able to do this.中文 · 翻译现在 EML 产能严重堵塞，CW 激光器也开始卡脖子了。 据我所知，$COHR 正从 $LITE 手里买 EML，因为他们自己产量不够。 $LITE 大部分产线都给了 EML，所以也没余力生产足够多的 CW 激光器。 于是他们（从财报电话会里透露）开始向竞争对手买 CW 激光器——可能是住友、古河等——而这些货源大概率最终流进 $NVDA 的订单里。 然后 $AMD 和你那些超大规模客户也急需产能，但我们早就供不应求了。 现在但凡有新产能上线，基本都会被抢光，因为看起来未来几年都会持续短缺（Lumentum 都已经卖到了 2028 年）。 这和 SK 海力士/三星/美光的存储芯片情况不太一样，但 EML 的护城河依然很深，能做的玩家可能一只手就数得过来。
- English · 原文Trendforce reports that $AMD is actively trying to secure CW laser supply with multiple major procurement orders... Is probably just the start of the bottleneck? There's not much independent capacity in Western supply chains left other than $SIVE or $AAOI and maybe Macom. Especially after Lumentum/Coherent got locked up with multi-year agreements with Nvidia. (disclosure, own Sive and aaoi) Lumentum is already CW laser constrained and is likely buying off Japanese companies like Sumitomo/Furukawa if I had to guess per ER, and those are probably running at max capacity. From the Trendforce report, this is: "to ensure that its future capacity will not be constrained by NVIDIA and other major Cloud Service Providers (CSPs)." I wouldn't be surprised if other hyperscalers like Amazon, Microsoft, and others saw Nvidia / AMD signing LTAs, and are trying to secure capacity next. A lot of it is game theory on not getting choked out by competitors, and looks like AMD is tipping the first domino after Nvidia. But my opinion is that this just goes and show how invaluable this CW laser chokepoint is and the companies are inside it.中文 · 翻译Trendforce 报道说 $AMD 正在积极争取 CW 激光供应，下了好几笔大采购订单…… 这可能只是瓶颈的开始吧？ 西方供应链中，除了 $SIVE 和 $AAOI，可能还有 Macom 之外，已经没多少独立的产能了。 尤其是在 Lumentum/Coherent 和英伟达签了好几年协议，产能都被锁住之后。（披露：持有 Sive 和 aaoi） Lumentum 的 CW 激光供应已经很紧张了，根据财报我猜他们可能正在从住友/古河电工这类日本公司那里采购，而这些公司大概也已经满负荷运转了。 根据 Trendforce 的报告，这是： “为了确保其未来的产能不会受到英伟达和其他主要云服务提供商（CSP）的限制。” 如果亚马逊、微软等其他超大规模企业看到英伟达/AMD 签署了长期协议，然后也开始争取产能，我一点也不会感到惊讶。 很大程度上这是一种博弈论，为了不被竞争对手卡脖子，看起来英伟达之后，AMD 正在推倒第一块多米诺骨牌。 但我的看法是，这恰好说明了 CW 激光这个“瓶颈”有多么宝贵，以及处于这个瓶颈中的公司有多重要。
- English · 原文I’m not sure why many folks are super bearish on my high conviction $AAOI long… Ever since $30, then on the way up to $170. (Yes I do think every bear is wrong, we’ll see who’s right). They have scarce laser capacity that $AMD and other hyperscalers are looking for. While the entire industry is bottlenecked by $NVDA. Along with a US transceiver supply chain for mass production of 800g/1.6T (management - largest in America). While demand far exceeds supply and while assembly gets outsourced to Asia. Then they’re quoting $471M monthly revenue in H1 entering H2 of 2027. Which is $5.6B ARR, off a $13.5B MC… While a lot of major inflection volume hits even later in 2028. As for fluctuations, there might be active $600M ATMs that get tapped into at random times. And random bear posts + macro from time to time that cause more volatility (eg. Analyst notes saying bear on $LITE due to false CPO delay rumors, then that brings down others in the sector). Also we’re a year out so timelines are still a little early. I haven’t seen such fast revenue ramp since $NBIS.中文 · 翻译不太理解为什么很多人对我的重仓多头 $AAOI 极度看空…… 从 $30 开始，一路涨到 $170 的时候都有人看空。（对，我就是觉得每个空头都错了，咱们看看最后谁对。） 他们有稀缺的激光产能，这正是 $AMD 和其他超大规模数据中心在找的。 而整个行业都被 $NVDA 卡着瓶颈。 加上他们有美国本土的收发器供应链，用于大规模生产 800g/1.6T（管理层表示——这是全美最大的）。 同时需求远超供给，而组装环节却外包到了亚洲。 然后他们给出的指引是，2027 年上半年到下半年过渡期间，月营收 4.71 亿美元。这相当于年化经常性收入 56 亿美元，而市值才 135 亿美元…… 而且到了 2028 年更晚的时候，还有大量重要的爆发量会出现。 至于波动，可能时不时会有 6 亿美元的活跃 ATM 增发被拿出来用。 再加上时不时冒出来的随机看空帖子和宏观因素，导致更大波动（比如某分析师报告因为假的 CPO 延迟传闻看空 $LITE，结果把板块内其他股票也带下来了）。 另外我们离那个时间点还有一年，所以时间线确实还有点早。 自打 $NBIS 之后，我就没见过这么快的营收爬坡。
- English · 原文Mizuho Research: No delays on CPO or 800v dc. Revised up optical engine projections from $NVDA demand ramp. - Next phase in CPO for long term, believes InP DFB lasers remains the focus (hello $SIVE). VSCEL and microLED remain "unproven" in short distance in rack + chip to chip for 1.6T+ - HVDC deployment on track, 800VDC incremental volume 2027, with higher penetration in 2028. What a stupid CPO related selloff recently.中文 · 翻译瑞穗研究：CPO 或 800V 直流电没有延迟。 由于 $NVDA 需求爬坡，上调了光学引擎预估。 - CPO 的下一阶段是长期方向，他们认为 InP DFB 激光器仍然是重点（你好啊 $SIVE）。VSCEL 和 microLED 在 1.6T+ 的机架内短距离 + 芯片到芯片场景中仍然"未经证实"。 - HVDC 部署按计划进行，800V 直流电增量产出在 2027 年，2028 年渗透率更高。 最近这波跟 CPO 有关的抛售真是蠢到家了。
- English · 原文Still holding Foci/Nextronics. There was a CPO bear post from an analyst that said there would be delays (which $NVDA denied), which caused CPO related stocks to crash. Personally think Foci will end up a bottleneck for $TSM COUPE / $NVDA and Nextronics to be a beneficiary of optical components relative to MC in $NVDA supply chains. We'll see how this plays out.中文 · 翻译还在拿着 Foci/Nextronics。之前有个分析师发了篇关于 CPO 的看空帖，说项目会有延迟（$NVDA 否认了这点），结果 CPO 相关股票就崩了。 我个人觉得，Foci 最终会成为 $TSM COUPE / $NVDA 的一个瓶颈，而 Nextronics 则会受益于 $NVDA 供应链里光学组件相比 MC 的地位。 看后面怎么走吧。
- English · 原文I did say $MU looked like the next $NVDA. Now we're at a $1.23T MC. Started talking more about Samsung Electronics/Sk Hynix back in 2025. Put more concentration into the memory theme like $SNDK and others, Jan of this year. And I'm glad my prediction with Micron + memory is playing out well! Hope people had fun with $EWY longs too, those are up a lot.中文 · 翻译我早就说过 $MU 看起来就是下一个 $NVDA。现在市值都到 1.23 万亿了。 从 2025 年那会儿就开始更多讨论三星电子 / SK 海力士了。 今年 1 月就更集中布局存储这个方向，比如 $SNDK 和其他几只。 我很开心美光 + 存储这条线的预测，行情正在应验！ 希望 $EWY 的多头仓位，大家也赚得开心，涨了很多。
- English · 原文Other way around, $NVDA bottlenecked the entire industry for EML capacity. And did the same with CW capacity ONCE AGAIN with $LITE, $COHR, and $MRVL (if they have LTA in place with Celestial) I said this a few months ago, we'd see this exact same playbook. But $AMD, $AMZN, $META, and others are just so stupidly slow that they let themselves get bottlenecked. Now there's only a few merchant players like $AAOI, $MTSI, and $SIVE that they all need to fight over.中文 · 翻译反过来看，是 $NVDA 卡住了整个行业在 EML 产能上的脖子。 而且它又干了一次同样的事情——在 CW 产能上，把 $LITE、$COHR 和 $MRVL 也给卡住了（前提是他们跟 Celestial 签了长期协议）。 我几个月前就说过，我们会看到完全一样的剧本。但 $AMD、$AMZN、$META 这些家伙动作慢得离谱，结果让自己也被卡在瓶颈里。 现在只剩少数几家商业玩家了，比如 $AAOI、$MTSI 和 $SIVE，大家都得争着抢。
- English · 原文New reports that $AMD is scrambling for CW laser supply. And is negotiating large-scale purchase orders for CW Lasers to ensure its production capacity is not constrained by $NVDA (Trendforce) Obvious CW laser beneficiaries: - $SIVE (AMD went to GFS for CPO, Sivers reference laser level) - $AAOI (Rosenblatt analyst checks) Lumentum/Coherent are kinda booked out way into 2028 as well. Lumentum is especially constrained for CW capacity already from existing EML contracts (so they probably are buying from Sumitomo/Furukawa and co). Maybe Macom and Japanese giants still have spare capacity. (disclosure, own aaoi/sivers). I predicted this last year and said hyperscalers should go more upstream to secure capacity... at laser levels, epiwafer levels, or even inp substrate levels. To not get bottlenecked by Nvidia.中文 · 翻译最新消息称 $AMD 正在疯狂寻找 CW 激光器供应。 并且正在洽谈大规模购买 CW 激光器的采购订单，以确保其产能不受 $NVDA 限制（Trendforce 消息） 明显的 CW 激光器受益方： - $SIVE（AMD 找 GFS 做 CPO，Sivers 是参考激光级别） - $AAOI（Rosenblatt 分析师的调研确认） Lumentum/Coherent 的产能基本已经排到 2028 年了。 Lumentum 的 CW 产能尤其紧张，因为已经签了很多 EML 合同（所以他们可能正在向 Sumitomo/Furukawa 等厂商采购）。 也许 Macom 和日本巨头还有多余产能。（声明：我持有 aaoi/sivers）。 我去年就预测过，说过超大规模云厂商应该更向上游去锁定产能……在激光器层级、外延片层级、甚至 InP 衬底层级。 这样才能不被 Nvidia 卡脖子。
- English · 原文$SIVE is the next SIVE. Don’t think you’ll find another company. That’s qualified and likely primary/sole source with: - $JBL and other pluggable hyperscaler suppliers - Ayar and the $NVDA NVLink CPO ecosystems While being the foundational reference laser for $GFS and pluggable/CPO/NPO deployments. That hyperscalers like $AMD and others use, at current valuations. Even $POET buys $SIVE lasers and Poet is about the same valuation just off having one $50m purchase agreement. Amount of hyperscaler suppliers for 2027 into 2028 is just ridiculous. From the general meeting today in a few hours, we’ll hopefully see NASDAQ listing timelines confirmed. So they can have room for M&A to TAM expansion and to make each laser they sell more valuable. Following what $LITE did to grow into a $75B company.中文 · 翻译$SIVE 就是下一个 $SIVE。 别指望能找到第二家这样的公司。 具备资质、并且很可能是主要/独家供应商，对接的是： - $JBL 以及其他可插拔超大规模客户供应商 - Ayar 和 $NVDA NVLink CPO 生态系统 同时还是 $GFS 以及可插拔/CPO/NPO 部署方案的基础参考激光器。 而且像 $AMD 和其他超大规模客户都在用，以当前的估值来看。 就连 $POET 也要买 $SIVE 的激光器，而 Poet 的估值跟 $SIVE 差不多，只不过靠的是那个 $5000 万的采购协议。 2027 到 2028 年的超大规模供应商数量简直离谱。 希望今天几个小时后的大会上，我们能听到纳斯达克上市时间表得到确认。 这样他们才有空间做并购来扩大 TAM（总可寻址市场），并让每一颗售出的激光器都更有价值。 就像当年 $LITE 那样，一步步成长为一家市值 750 亿美元的公司。
- English · 原文$NVDA and $GOOGL lead 800V DC ahead of schedule. "Ahead of schedule", pulled up to Q3 2026 with small volume shipments starting . - Delta Electronics (2308), $VRT - Song Chuan Precision (7788) - Schneider Electric, Eaton, Siemens. All flagged as beneficiaries. "Market sources indicate that Nvidia’s Vera Rubin platform and Google’s next-generation AI data centers will be the first to adopt the technology" Source: Commercial Times The power semi trade should be happy to hear this.中文 · 翻译$NVDA 和 $GOOGL 提前引领 800V 直流电。 "提前于原计划"，已提前至 2026 年 Q3，小批量出货即将开始。 - 台达电子 (2308)，$VRT - 松川精密 (7788) - 施耐德电气、伊顿、西门子。 全都被列为受益方。 "市场消息指出，Nvidia 的 Vera Rubin 平台和 Google 的下一代 AI 数据中心将率先采用这项技术" 来源：工商时报 做功率半导体的应该听了会开心。
- English · 原文If you haven’t noticed too with my other investment themes with 800V DC and CPO recently. It’s investing in $NVDA, America’s national champion in AI, and securing their supply chains. Many things feel technologically difficult with yields to substrate supply. People can always bear post laser capacity or export control bottlenecks and tell people to short Nvidia’s supply chains due to difficulties. But by investing in the critical companies to give them more capex spend for FAU capacity / yields. Or funding upstream red phosphorus/InP substrate capacity or SiC/GaN capacity. It builds up Western supply chains to make what’s technologically challenging, possible. Also I believe in Jensen.中文 · 翻译如果你还没注意到我最近其他投资主题，比如 800V 直流快充和 CPO 的布局。 其实就是投资 $NVDA——美国在 AI 领域的国家冠军——并保障其供应链的稳定。 从良率到衬底供应，很多环节在技术上都很难。 总有人能拿后道激光产能或出口管制的瓶颈来做空，告诉别人“英伟达的供应链太难了，做空吧”。 但如果你去投资那些关键公司，给它们更多资本开支来提升 FAU 产能和良率。 或者去资助上游的红磷/InP 衬底产能、SiC/GaN 产能。 就能建立起西方的供应链，让那些技术上很困难的事情变得可行。 而且，我相信黄仁勋。
- English · 原文@ratna555 I see the Lightmatter/Ayar type companies, probably going higher than $5B if they IPO. Since they're both part of $NVDA NVLink CPO ecosystem, heavy backers like NVidia/Intel/AMD/Google, and popular theme.中文 · 翻译@ratna555 我看好那些 Lightmatter/Ayar 类型的公司，如果它们 IPO 的话，估值很可能超过 50 亿美元。 毕竟它们都属于 $NVDA NVLink CPO 生态系统的一部分，背后有像 NVIDIA/Intel/AMD/Google 这样的重量级投资方，又是当下热门的题材。
- English · 原文Why do I keep getting these questions!!! $XFAB is building a Silicon Photonics foundry alternative to $TSEM and $GFS. And has Europe backing it + $NVDA evaluations. It takes time... Like October 2026, should finish up development. Then 2027 production scaled into 2028 (mass production), since they've been working on it since 2023. Everyone thinks it’s a depressed automaker supplier right now. And thankfully with European names they tend to look at TTM revenue over forward growth. So somehow it’s leading EU’s efforts to create a $TSEM silicon photonics foundry + supply chain at ~€1.1B MC? That R&D directors from ASE cite + others as future CPO routes. I’m might just really early to a lot of things, but of course most of the risk/reward comes from taking a little leap of faith in seeing it commercialized. Otherwise people can take the de-risked route with Tower directly (which I also wrote a thesis on awhile back and also like).中文 · 翻译为什么我一直被问到这些问题！！！ $XFAB 正在打造一个硅光子学代工厂，作为 $TSEM 和 $GFS 的替代品。而且有欧洲支持它，再加上 $NVDA 的评估。 这需要时间……比如到 2026 年 10 月，开发应该能完成。 然后 2027 年产量会逐步扩大，到 2028 年进入（大规模量产），毕竟他们从 2023 年就开始搞了。 现在所有人都觉得它就是个低迷的汽车供应商。好在欧洲公司嘛，大家习惯看 TTM 营收，而不是盯着未来增长。 所以它就这么莫名其妙地领导了欧盟打造一个 $TSEM 式的硅光子学代工厂及供应链，市值大约 €11 亿？ 日月光（ASE）的研发总监和其他人都把它列为未来的 CPO 路线。 我可能在很多事情上都太早了，但当然大部分的风险回报就来自一点信念，赌它能商业化。 不然大家也可以直接选 Tower 这条更稳妥的路（我之前也写过一篇关于它的分析，也挺喜欢的）。
- English · 原文Basically this… and it’s how cycles work. Retail was early and completely frontran institutions on next architectural shifts. There was close to 0 US institutional ownership on $SIVE. And now you see active institutions like JP Morgan, Fidelity Research, and others on the cap table. Happened last year with $NBIS. > I called out close to <30% institutional accumulation and said they wanted more shares. > institutions bought up majority of the float > bunch of negative articles back then, now it’s positive and ATHs. Two years before it was $RKLB > Was long at $16, but institutional analysts kept giving record low PTs and told retail to sell, although it had such a high reusable rocket rate. > retail sold, institutional ownership stocked up > now it’s ATHs I expect Foci (3363) to be a bottleneck for both $NVDA and $TSM optical programs and now there’s firms implying you to sell that at $2.5B valuations alongside $HIMX. So if you see negative sellside reports or an uncanny wave of negative news, if’s a good signal they need liquidity. Recently some smaller hedge funds have been so desperate that they’re likely even using bot farms on X that told retail to sell lol… which I’ve uncovered recently. Regardless, it’s also why I spend a lot of time doing research on individual names so people can build their own conviction in the face of noise. Unfortunately, it’s just a part of life how the modern liquidity cycles/transfers of US retail -> Institutions work. They don’t work in the best interest of retail investors.中文 · 翻译基本就是这样……这就是周期的运作方式。 散户入场早，完全抢先了机构，抓住了下一波架构转型。 当时 $SIVE 几乎没有美国机构持仓。 而现在你看到摩根大通、富达研究部等一些活跃机构都出现在了股东名单上。 去年 $NBIS 也发生了同样的事。 > 我
- English · 原文@oops1GSP Bro everything has technical obstacles. I can bear post about HBM4 yields or glass substrate yields and basically everything too. $NVDA is a $5T company, they’re not $ASTS. I’m sure they know timelines and difficulties, so projections should be accurate.中文 · 翻译@oops1GSP 兄弟，所有东西都有技术难题。我也可以写一篇关于 HBM4 良率或玻璃基板良率的帖子，基本啥都能聊。 $NVDA 是家 5 万亿市值的公司，可不是 $ASTS。 我确信他们清楚时间线和难点，所以预测应该挺准的。
- English · 原文@lumingxi2025 No, $NVDA denied reports about 800v and CPO delays. I think I’ll trust Nvidia… since they probably have an idea on their own timelines…中文 · 翻译@lumingxi2025 不，$NVDA 否认了关于 800v 和 CPO 延迟的报道。 我觉得我会相信 NVIDIA……毕竟他们对自己时间表应该心里有数……
- English · 原文Morgan Stanley: $NVDA has denied the reports 800V DC has been pushed back. Recent SemiAnalysis reports run contrary to our own checks at Computex. Bro this has gotta be the dumbest CPO/800V selloff I’ve seen. Since the selloff from their claim $MU had 0 share of Nvidia HBM4 https://t.co/YX9apQSVLT中文 · 翻译摩根士丹利：$NVDA 否认了关于 800V DC 被推迟的报道。 SemiAnalysis 最近的报告与我们在 Computex 上的核查相反。 兄弟，这绝对是我见过的最蠢的 CPO/800V 抛售了。 自从他们宣称 $MU 拿不到任何英伟达 HBM4 份额后，市场就开始抛售 https://t.co/YX9apQSVLT
- English · 原文If you want a TLDR of today: > be $NVDA, $5T company. Force shift to 800V DC and CPO > analyst: I don’t think u can do it in time! > market: “I don’t trust Nvidia, time to sell everything” > Nvidia and Lumentum executives after: Bullish on CPO, timelines accelerating. ???中文 · 翻译如果你想要今天的一个 TLDR： > 你是 $NVDA，一家市值 $5T 的公司。强行转向 800V 直流电和 CPO > 分析师：我觉得你来不及搞定！ > 市场：“我不相信英伟达，是时候清仓了” > 英伟达和 Lumentum 的高管随后：看好 CPO，时间表在加速。 ???
- English · 原文@xnoahwang Imagine being $NVDA, the most powerful company in the world, with high visibility of their own timelines, saying there’s no delays. Then external analysts go and say every architecture Nvidia is doing is gonna be delayed by awhile. Yeah… I’m going long with Nvidia here.中文 · 翻译@xnoahwang 想象一下，你是个 $NVDA，全世界最牛的公司，对自己的时间表看得清清楚楚，然后说没有延迟。 结果外部分析师跑出来说，Nvidia 做的每个架构都要推迟一阵子。 行吧…… 我这边还是做多 Nvidia。
- English · 原文$LITE Management Speech from Mizuho Technology at today’s conference. The company expects to start shipping CPO scale up optical products in the second half of 2027. With formal ramp up in 2028. No delays, as this aligns with previous timelines shared. So today, we also got confirmation from $NVDA SVP no delays on CPO scale out timelines H2 2026, and they’re beginning mass production. And $LITE management also stated no delays on CPO scale up timeline. The leading companies in Nvidia and Lumentum probably know their own timelines the better than incorrect analyst reports telling them no. And both are incredibly bullish on TAM and opportunities.中文 · 翻译今天瑞穗科技在会议上听了 $LITE 管理层的发言。 公司预计 2027 年下半年开始出货 CPO 规模化光学产品。 2028 年正式上量。 没有延期，这与之前分享的时间表一致。 所以今天，我们也得到了 $NVDA 高级副总裁的确认， CPO 规模化时间表没有延期，仍为 2026 年下半年，而且他们已经开始量产。 $LITE 管理层也表示，CPO 规模化时间表没有延期。 英伟达和 Lumentum 这两家领头羊，大概率比那些瞎说"不可能"的分析师报告更清楚自己的时间表。 而且两家都对总可用市场和机会非常看好。
- English · 原文$NVDA Networking Senior Vice President refuting recent analyst reports on delays: - “ the most exciting stuff is co-packaged optics.” - There is no delay in H2 CPO product delivery schedule. - CPO switch will enter mass production and begin ramping up customer deliveries as planned in the second half of 2026 This was a media article, original interview source credit should have been credited to Tae Kim / Computex. Something fun to note too was this quote “Gilad was VERY enthusiastic about the CPO ramp from Nvidia.” Both near term and long term. Yeah… I’m extremely bullish on CPO alongside Nvidia.中文 · 翻译$NVDA 网络部门高级副总裁驳斥近期分析师关于延迟的报告： - “最激动人心的东西是共封装光学。” - H2 CPO 产品交付时间表没有延迟。 - CPO 交换机将按计划在 2026 年下半年进入量产并开始向客户交付。 这是一篇媒体报道，原始采访来源应该归功于 Tae Kim / Computex。 还有个有趣的细节，这句引语是：“Gilad 对 Nvidia 的 CPO 上量表现得极为兴奋。” 无论是短期还是长期。 是的……我极度看多 Nvidia 的 CPO。
- English · 原文I’m just sharing what’s imminent with architectural shifts pushed by $NVDA. Retail managed to completely frontrun institutions on multiple names, but institutions need liquidity to enter. It’s just interesting how every major industry player confirms the same timelines, then a questionable analyst firm that said $MU had no HBM4 share could write a hit piece then be completely incorrect again.中文 · 翻译我只是在分享 $NVDA 推动架构变革下即将发生的事。 零售交易者成功地在好几个标的上完全抢跑了机构，但机构需要流动性才能进场。 有趣的是，每个行业主要玩家都确认了相同的时间线，然后一家曾声称 $MU 没有 HBM4 份额的可疑分析机构，就能写一篇黑稿，结果再次完全错误。
- English · 原文@NathanJoooo This was in response to that erroneous article. I would trust Foxconn/Lumentum/Nvidia industry projections over an analyst firm that messed up $MU HBM4 so badly and said “no” to all their timelines together中文 · 翻译@NathanJoooo 这是针对那篇错误报道的回应。 我更相信富士康/Lumentum/英伟达的行业预测，而不是一家把 $MU HBM4 搞得一塌糊涂、并对所有时间线都说了“不”的调研机构。
- English · 原文CPO scale out earlier than expected: > Foxconn: est. units register upward and optical switches shipped early to $NVDA CPO scale up timelines from $LITE Mizuho Technology Conference today: “The company expects to start shipping Scale-Up optical products in the second half of 2027, with formal volume ramp-up in 2028” SVP $NVDA networking: “We’re going to ramp up CPO second half of this year”. No delay indications. I’m gonna go ahead and trust industry projections. Where they all reiterate faster timelines for scale out CPO H2 onward. And scale up CPO H2 2027 onward (with main growth happening 2028) Over a questionable motive analyst firm that said $MU had no share of HBM4 Rubin (causing a selloff) Where micron went out shortly later to into enter mass production. (Triple digit return shortly after) I think people going long on temporary bridge architectures from this incorrect report won’t be too happy. Appreciate the buying opportunity though.中文 · 翻译CPO 向外扩展时间表比预期提前： > 鸿海：预估单位出货量上修，光交换器提前出货给 $NVDA 今天 $LITE 瑞穗科技大会上的 CPO 向上扩展时间表： “公司预计在 2027 年下半年开始出货向上扩展的光学产品，2028 年正式放量” $NVDA 网络部门 SVP：“我们今年下半年就要开始上量 CPO”。没有任何延期的迹象。 我选择相信行业预测。他们都一致强调：向外扩展的 CPO 下半年起加速；向上扩展的 CPO 从 2027 下半年开始（主要增长在 2028）。 而不是相信动机可疑的分析机构——那家说 $MU 在 HBM4 Rubin 上没有份额（引发了一轮抛售）。 结果美光没过多久就宣布要进入量产（之后不久就翻了好几倍）。 我觉得那些因为这份错误报告而做多临时桥接架构的人，怕是不会太开心。 不过还是要感谢这个买入机会。
- English · 原文@QGrowthCap I mean $NVDA CEO did say memory would be a multi year shortage. With the way things are going with the nand/dram hikes, all your hyperscalers r gonna be in debt lol中文 · 翻译@QGrowthCap 我是说 $NVDA 的CEO确实说过，内存会短缺好几年。 就现在nand/dram涨价这架势来看，你那几家超大规模数据中心客户全都得负债累累哈哈。
- English · 原文Many work against the interest of retail investors. Especially as JP Morgan / US institutions are trying to buy up the float. You have a company funded with CHIPS Act, in major hyperscaler supply chains from $NVDA to others, powering Ayar for CPO, $JBL for pluggables. And now the reference laser for $GFS. With closest ramp with $AEVA / $POET likely coming next. And following that likely US NASDAQ Listing. Sivers is one of the most compelling photonics companies, and pretty sure all the deals established that already.中文 · 翻译很多动作都对散户投资者不利。 尤其是当摩根大通 / 美国机构正试图买断流通股的时候。 你有一家靠《芯片法案》资金支持的公司，身处从 $NVDA 到其他巨头的主要超大规模数据中心供应链中，为 Ayar 提供 CPO 方案，为 $JBL 提供可插拔模块。现在又成了 $GFS 的参考激光器供应商。 接下来最接近放量的很可能是 $AEVA / $POET。再之后，大概率就会在美国纳斯达克上市。 Sivers 是光子学领域最引人注目的公司之一，我相当确定已有的所有交易都已经证明了这一点。
- English · 原文@LightLogix I mean... > $SIVE reference laser for $GFS > $SIVE laser 1.6T LRO with $JBL + many more pluggable players > $SIVE + Ayar + $NVDA NVLink > likely $SIVE + Lightmatter / $MRVL Celestial with $NVDA NVLink > Likely $SIVE -> $AMD CPO ecosystem too. very compelling long idea.中文 · 翻译@LightLogix 我的意思是…… > $SIVE 是 $GFS 的参考激光器 > $SIVE 提供 1.6T LRO 激光方案，配合 $JBL 以及众多可插拔厂商 > $SIVE + Ayar + $NVDA NVLink > 很可能是 $SIVE + Lightmatter / $MRVL Celestial 搭配 $NVDA NVLink > 大概率 $SIVE 也会进入 $AMD 的 CPO 生态系统 这是个非常吸引人的做多逻辑。
- English · 原文@ark_btc I compiled 32+ different names related to $NVDA 800V DC mentioned by followers. And posted them in a simple format for people to do research on. Not my recommendations lol.中文 · 翻译@ark_btc 我把粉丝们提到的 $NVDA 800V DC 相关的 32 个以上不同名字都整理出来了。然后用一个简单格式发出来，方便大家自己去研究。 不是我的推荐哈，哈哈。
- English · 原文On top: $NVDA CEO also called out Silicon Photonics (optical networking) with memory. Stating that Nvidia would require “supply volumes beyond imagination”. What a bullish read through on the SiPH supply chain from $SIVE (now upstream Nvidia ecosystem) to $SOI https://t.co/m6jub4nfzx中文 · 翻译关键点：$NVDA 的CEO还特别提到了硅光子（光网络）与存储结合。 并表示英伟达将需要“超乎想象的供应量”。 这对于硅光子供应链来说，从$SIVE（如今已是英伟达上游生态系统的一部分）到$SOI，是一个多么看好的信号啊。 https://t.co/m6jub4nfzx
- English · 原文Oh look… $NVDA CEO warned memory shortage is expected to persist for many years, due to massive scaling demand of AI infrastructure. With further announcements tomorrow. $MU and $EWY (Samsung/SK Hynix) operating profit projections aren’t looking too crazy anymore? https://t.co/OvjyrifRtO中文 · 翻译哦你看……$NVDA 的 CEO 警告说，由于 AI 基础设施的规模需求激增，内存短缺预计会持续很多年。 明天还会有更多公告。 $MU 和 $EWY（三星 / SK 海力士）的营业利润预期，现在看起来也没那么夸张了吧？https://t.co/OvjyrifRtO
- English · 原文Sure, #1 thing is toxic financing structure/float dynamics. Best example is current Neoclouds landscape: - $IREN is basically trash, since they have $6,000,000,000 ATMs and virtually infinite dilution, likely selling into every rally (structural overhang) - While $NBIS is now YTD 153%+, from optimal structures (eg. $NVDA direct funding, mix of convertibles, etc.). - On the other hand, $CRWV has endless debt interest given they took out high interest rate loans to finance GPUs. It's extremely nuanced, but you need to take a look at the float dynamics. If they're legitimately a good company, then it might be a good idea to go long after all the existing holders get diluted to oblivion. But if you care about your equity appreciation, it's a good idea to stay far away from toxic financing structures or toxic overhang (eg. debt interest, that eats away at a company FCF long term) With smaller companies, they have this all the time, like $SLNH, where there's new $500m ATMs on a $250m MC. Or like $BKKT where there's endless dilution to fund executive pay. With these companies you're basically transferring your money over to the company while influencers talk about them. So those are red flags. With many software names like $SNAP, they mask stock-based compensation with profitability. So while the company optically looks profitable, you'll likely see the value of your equity decrease due to dilution. There's endless types of these share structures you need to look when screening ideas.中文 · 翻译当然，第一重要的是毒性融资结构 / 流通股动态。 最好的例子就是当前的新一代云服务商格局： - $IREN 基本就是垃圾，因为他们有 60 亿美元的 ATM 增发额度，几乎无限稀释，大概率每次反弹都在卖（结构性压制） - 而 $NBIS 今年至今已经涨了 153%+，靠的是最优结构（比如 $NVDA 直接投资、混搭可转债等）。 - 另一方面，$CRWV 因为用高利率贷款买 GPU，背负着无尽的债务利息。 这事非常微妙，但你必须关注流通股的动态。 如果这家公司确实质地不错，那等现有持有人被稀释到快没了之后，或许是个做多的好时机。 但如果你在意自己股权的增值，就最好离这种毒性融资结构或毒性压制（比如债务利息长期蚕食公司自由现金流）远远的。 小公司里这种情况比比皆是，比如 $SLNH，市值才 2.5 亿美元，却搞了个新的 5 亿美元 ATM 增发。 或者像 $BKKT，为了给高管发工资搞了没完没了的稀释。 这类公司，基本上就是 KOL 们在吹的时候，你把钱转给了公司。所以这些都是红旗。 很多软件股，比如 $SNAP，会用盈利数据来掩盖股权激励。所以虽然公司账面看着是赚钱的，但你的股权价值很可能因为稀释而缩水。 筛选标的时，这类股权结构无穷无尽，你必须仔细看。
- English · 原文@Jornka329996 > posts an idea about a 2027 silicon photonics foundry evaluted by $NOK and $NVDA last week > sells on macro drop few days after > complain中文 · 翻译@Jornka329996 > 上周发了个关于2027年硅光子晶圆厂的想法，还得到了 $NOK 和 $NVDA 的评估 > 过了几天因为宏观下跌就卖掉了 > 然后开始抱怨
- English · 原文hmm, i prefer all your upstream chokepoints over $NVDA long term since those will be re-rated the most (nvidia already largest company in the world) pretty sure hyperscaler ASICs would eventually siphon off $NVDA demand like $GOOGL TPU, $AMZN trainium programs. wouldn't be too positive for expontentially compounding revenue growth since hyperscalers were Nvidia's original main revenue stream (even indirect via Neoclouds). But $NVDA's kinda stalling everyone elses buildout by bottlenecking their programs eg. EML/laser capacity agreements years out too. And took stakes in $MRVL / $LITE / $COHR / $INTC etc. making them adopt to $NVDA standards or just owning a large %. So even if they're delaying other programs + their biggest growth vector kinda falls off one day, like how things are shifting already shifting to ASICs for inference. They'll still probably be fine given ownership stakes + will serve companies/countries outside of hyperscaler cash cows (just less revenue)+ made so much before then. But that's probably why p/e keeps going down despite revenues going up, since idk if markets thinks that growth will last forever. Or could be totally wrong and they just keep leapfrogging generation by generation + AI pie keeps growing with Jensen's 4T 2030 capex number.中文 · 翻译嗯，长期来看，我更看好你提到的所有上游瓶颈环节，而不是 $NVDA，因为那些东西的估值重估空间最大（英伟达已经是全球最大公司了）。 我相当确信，超大规模云的 ASIC 最终会分流 $NVDA 的需求，就像 $GOOGL 的 TPU、$AMZN 的 Trainium 项目那样。这对指数级复利的营收增长可不太有利，因为超大规模云本来就是英伟达最初的主要收入来源（甚至通过 Neocloud 这种间接渠道也是）。 不过 $NVDA 也在卡别人脖子，拖慢大家的建设进度，比如把 EML/激光产能协议签到了好几年后。而且它还入股了 $MRVL / $LITE / $COHR / $INTC 等公司，要么让它们采用 $NVDA 的标准，要么直接拿大头股份。 所以，哪怕它拖慢了其他项目 + 自己最大的增长引擎某天会熄火（就像现在推理已经在往 ASIC 转移了）， 考虑到它手里那一堆股权 + 还能继续服务超大规模现金牛之外的公司和国家（虽然收入会少一些）+ 之前早就赚得盆满钵满，它大概率还是能过得不错。 不过，这大概也是为什么它收入一直在涨、市盈率却持续走低的原因——大概市场也不确定这种增长能不能永远持续。 当然，也可能我完全错了，它们就是一代代持续领先 + AI 蛋糕持续做大，配合 Jensen 说的 2030 年 4 万亿 capex 目标。
- English · 原文Fun times with market corrections. Leaders from $NVDA down -4.87% to $MU down -7.03%. High beta names like $PL down -22.02%. Funny to see media always trying to explain like: "Micron suffers record wipeout as Broadcom casts a shadow over chip stocks " Broadcom projected insatiable demand into 2028, just made up narratives. Nothing's changed the AI buildout aside from increasing capex. Main material thing was rate hike probabilities increase. But you have random ones like these few times a year into ATHs. Personally wouldn't try and trade fed decision probabilities and stay long on current company projections (eg. $AAOI $471m h1 2027)中文 · 翻译市场回调的时候还挺有意思的。 从 $NVDA 跌 4.87% 到 $MU 跌 7.03%。高贝塔票像 $PL 直接跌了 22.02%。 挺好笑的，媒体总在那拼命解释，比如： "博通给芯片股蒙上阴影，美光遭遇创纪录暴跌" 博通预测需求旺盛到 2028 年，纯粹是编故事。AI 基建这边除了资本支出在增加，什么都没变。 真正实质性的变化是加息概率上升了。 但每年总有那么几次，在大涨新高之后冒出这类随机回调。 我个人不会去押注美联储决策的概率，继续拿着现有公司指引的仓位（比如 $AAOI 2027 上半年 4.71 亿美元）。
- English · 原文$SIVE is #1, $AAOI is #2 used wrong wording above. Generally a fan of: - $SIVE (CPO lasers) - $AAOI (End-to-End pluggable/cpo) - Foci (FAU +passive components $TSM COUPE / $NVDA) - Shunsin (Packaging/Test) - Win Semi (foundry) - $TSEM (foundry) - $SOI (silicon photonics) - Nextronics (CPO connector / cage thermal module) And a few others.中文 · 翻译$SIVE 是 #1，$AAOI 是 #2——刚才用词有误。 整体上我比较喜欢： - $SIVE（CPO 激光器） - $AAOI（端到端可插拔 / CPO） - Foci（FAU + 无源器件 + $TSM COUPE / $NVDA） - Shunsin（封装 / 测试） - Win Semi（晶圆代工） - $TSEM（晶圆代工） - $SOI（硅光子） - Nextronics（CPO 连接器 / 散热模块） 还有几家别的。
- English · 原文H1 2027 all the 1.6T pluggable players like $JBL. Maybe Innolight/Eoptolink and other pluggable players are added too. We'll find out soon since $SIVE said there were undisclosed pluggable players they're working with. H2 2027 for all the main CPO scale up applications from $NVDA NVLink CPO ecosystem players like Ayar. Markets are forward looking ~8M in advance usually.中文 · 翻译2027年上半年，所有1.6T可插拔的玩家，比如$JBL。 可能Innolight/Eoptolink还有其他可插拔的公司也会加进来。我们很快就会知道，因为$SIVE提到他们还有些没公开的可插拔合作伙伴在谈。 2027年下半年，$NVDA的NVLink CPO生态系统里的玩家（比如Ayar）将全面铺开CPO规模化应用。 市场通常提前大约8个月定价。
- English · 原文@Jornka329996 Are you high, I posted about $XFAB this week. Their silicon photonics platform with $NVDA and $NOK scales h2 2027 / 2028. I think it’s a heavily derisked precommercial long that looks like the next $TSEM. With upside from SiC/GaN. Just needs time.中文 · 翻译@Jornka329996 你是不是上头了，我这周发过 $XFAB 的内容啊。 他们跟 $NVDA 和 $NOK 合作的硅光子平台，量产要到 2027/2028 下半年。 我觉得这是一个风险大幅释放的上市前长线标的，看起来就像下一个 $TSEM。还有 SiC/GaN 带来的上涨空间。 就是需要时间。
- English · 原文$AAOI is one of the names I keep averaging up on since $28. Just from random shower thoughts… I feel like it’s just imminent to double or triple if they execute? There’s just too much demand for 800g/1.6T optical transceivers… Then this company is targeting the largest capacity in the US, with extreme vertical integration. I think something to keep in mind is sovereign DCs / T2 AI DCs which increase the demand for 800g as hyperscalers upgrade to 1.6T. So demand for 800g can actually keep increasing… Then there’s the analyst rumors of $AAOI conversations with $AMD / $NVDA. Which is kinda expected given everyone is getting their capacity allocated way into 2028. Nvidia always starts first and causes bottlenecks for everyone else as seen with EML, so not surprising if another hyperscaler learned their lesson this time? Also, everyone seems to be modeling lower ASP at scale. But if this ends up a major bottleneck H1 next year as expected… Could see unexpected price hikes + margin expansion across the board from $AAOI, $LITE, and others not really modeled in.中文 · 翻译$AAOI 是我从$28就开始持续加仓的标的之一。 就是随便洗澡时瞎想……感觉如果他们执行到位，翻倍甚至翻三倍几乎是迟早的事？ 毕竟800G/1.6T光模块的需求实在太大了…… 而这公司瞄准的是美国最大产能，还搞了极端的垂直整合。 我觉得有一点值得记住——主权数据中心 / Tier 2 AI 数据中心也会增加800G的需求，因为超大规模云厂商正在往1.6T升级。 所以800G的需求实际上会持续增长…… 然后还有分析师传言说$AAOI在和$AMD / $NVDA洽谈。考虑到所有人都在抢产能排期到2028年，这其实挺符合预期的。 英伟达总是先动手，然后像EML那样卡住其他人的瓶颈，所以如果这次有另一家超大规模云厂商吸取了教训，也不奇怪吧？ 另外，似乎所有人都在按规模量产后ASP下降来做模型。但如果明年上半年真的像预期那样成为主要瓶颈…… 那可能就会出现意料之外的涨价 + 利润率全面扩张，$AAOI、$LITE 以及其他标的都还没把这点计入模型。
- English · 原文This timeline keeps getting more and more unreal… $NVDA Jensen Huang to meet Faker (League of Legends) I did talk with Jensen few years back about PUBG and GPU shortages for mining. And he was enthusiastic about ideas for cross-platform gaming + getting enough allocation to gamers. Fast forward to now, feels like Nvidia kinda ignored gamers… for AI, which is understandable. So this is a nice symbolic message for a return to Nvidia’s roots.中文 · 翻译这条时间线真是越来越离谱了…… $NVDA 黄仁勋要和 Faker（《英雄联盟》选手）见面 几年前我还跟黄仁勋聊过《绝地求生》和挖矿导致的 GPU 短缺问题。 当时他对跨平台游戏的想法特别来劲，还说要给玩家多分配些货。 现在再看，感觉 Nvidia 有点把玩家晾一边了……都去搞 AI，当然也能理解。 所以这次会面算是个不错的象征性信号，表示要回归 Nvidia 的老本行了。
- English · 原文I’ve repeated this for every single stock, but insider selling means literally nothing. A managing director of wireless transferring ownership over to US investors doesn’t change any fundamentals about the business. CEO of $NVDA and $MU sell shares all the time but nobody bats an eye for a reason中文 · 翻译这一条我对每支股票都反复强调过：内部人减持根本说明不了任何问题。 一位无线业务董事总经理把股份转给美国投资者，不会改变这家公司的基本面哪怕一丁点。 $NVDA 和 $MU 的 CEO 也一直在卖股票，但大家从来不眨一下眼，这背后是有原因的。
- English · 原文$SIVE looks like both a chokepoint and a bottleneck for CPO next year. Keep seeing information published from nontechnical people who miss any nuances. Here’s the reason why: 1. CW lasers are bottlenecked signaled by $LITE earnings. Laser fabs are heavily allocated to EML likely from former $NVDA contracts. -> Sumitomo/Furukawa = bottleneck -> Win Semi = bottleneck $SIVE does fab-lite, so are they a bottleneck? Yes, $SIVE sits in the laser bottleneck since control output supply of CW lasers from Win Semi and other fabs from allocation way early on (CEO stated they working with more capacity from other players as well). Perfect example is Kioxia/Sandisk. $SNDK controls NAND output, so they’re a bottleneck because they control final pricing. Demand exceeding supply from Ayar, Jabil, other pluggable vendors + Nvidia NVLink CPO ecosystem… final laser supply owned by $SIVE makes Sivers a bottleneck. $SIVE is also likely primary/sole source for Jabil, Gen-1 Ayar, $MRVL Celestial, and other hyperscaler asic/merchant CPO routes. So no way to get around it (can’t hot-swap single channel cw lasers with Sivers) 2. $SIVE is a chokepoint over CPO. $NVDA use $COHR, $LITE (which likely sources external cw capacity from Japanese competitors) $AVGO is likely vertically integrated as well. However: the entire ecosystem around it from ASIC programs (Marvell, AlChip, etc) and merchant programs (Ayar, Lightmatter, Lightelligence) Are all likely designed around $SIVE. Ayar for example, likely tried to multi-source with $MTSI / $LITE back in 2022 but their lasers probably couldn’t match the level of Sivers specification with arrays (removed Lumentum / Macom from their supply chain site recently) If there’s no alternative at least for the initial generations (obviously they’re working to multi-source). That makes $SIVE a structural chokepoint to go through for lasers. Even if you look at the 1.6T LRO $JBL designed, they achieved a “drastic moat” with performance built around $SIVE likely sole source. $SIVE is also the foundry level reference laser design for $GFS, which your hyperscalers use like $AMD (likely using Sivers + maybe Ayar for gen1): If every major player, who hasn’t achieved vertical integration (Nvidia/Broadcom) is using Sivers for CPO… That makes them a chokepoint. Just look at the entire CPO $NVDA NVLink ecosystem partners: every single one are all likely using Sivers. And they all use $GFS as well (where Sivers is default reference). So $SIVE is both a chokepoint and bottleneck when CPO really scales up H2 2027, over one of the biggest architectural shifts of all time (near $0 -> $81B or $91B TAM in the next 1 1/2 years from GS research note) This is why I say $SIVE looks like it could be the next $75B $LITE over the next couple years. All of this should play out next year. And it’s still trading less than a company with $50M in purchase agreements that buys Sivers lasers to repackage them.中文 · 翻译$SIVE 看起来既是明年 CPO 的关卡，又是它的瓶颈。 总能看到非技术人员发布的信息，完全忽略了各种细节。 下面说说原因： 1. CW 激光器受到 $LITE 财报所显示的瓶颈制约。 激光晶圆厂的大量产能都被分配给了 EML，很可能来自 $NVDA 之前的合同。 → 住友电工/古河电气 = 瓶颈 → Win Semi = 瓶颈 $SIVE 做的是 Fab-lite 模式，那它算不算瓶颈？ 是的，$SIVE 就卡在这个激光瓶颈里，因为他们很早就掌控了从 Win Semi 和其他晶圆厂流出的 CW 激光器产能分配权（CEO 说过他们也在与其他厂商合作，增加更多产能）。 完美的例子是 Kioxia/Sandisk。$SNDK 控制着 NAND 的产出，所以他们就是瓶颈，因为他们控制着最终定价。 Ayar、Jabil、其他可插拔模块厂商，加上 Nvidia NVLink CPO 生态，需求超过了供给……最终激光器供应掌握在 $SIVE 手里，这就让 Sivers 成了瓶颈。 $SIVE 很可能也是 Jabil、Gen-1 Ayar、$MRVL Celestial，以及其他超大规模 ASIC/商用 CPO 路线的主供或唯一来源。所以绕不开它（单通道 CW 激光器没法热替换掉 Sivers）。 2. $SIVE 是 CPO 的关卡。 $NVDA 用的是 $COHR、$LITE（后者很可能从日本竞争对手那里外购 CW 产能）。 $AVGO 很可能也是垂直整合的。 然而：它周围整个生态，从 ASIC 项目（Marvell、AlChip 等）到商用项目（Ayar、Lightmatter、Lightelligence），很可能都是围绕 $SIVE 设计的。 比如 Ayar，2022 年可能尝试过从 $MTSI / $LITE 做多源采购，但它们的激光器在阵列规格上可能达不到 Sivers 的水平（最近他们从供应链网站上去掉了 Lumentum / Macom）。 如果至少在初期几代产品中没有替代方案（显然他们正在努力做多源采购），那就让 $SIVE 成了激光器领域一个结构性的、必须通过的关卡。 即使看看 $JBL 设计的 1.6T LRO，他们也通过围绕 $SIVE（可能是唯一来源）的性能，建立了一道“巨大的护城河”。 $SIVE 也是 $GFS 的代工厂级参考激光器设计，而你的超大规模客户（比如 $AMD，Gen1 很可能用的是 Sivers + 也许还有 Ayar）都在用 $GFS： 如果每一个尚未实现垂直整合的主要玩家（Nvidia/Broadcom）都在 CPO 中使用 Sivers…… 那他们就构成了一个关卡。 看看整个 $NVDA NVLink CPO 生态的合作伙伴：每一个都很可能在使用 Sivers。而且它们全都用 $GFS（Sivers 在这里是默认参考设计）。 所以当 CPO 在 2027 年下半年真正起量时，$SIVE 既是关卡又是瓶颈，而这次 CPO 是有史以来最大的架构变革之一（根据高盛研究报告，TAM 从接近 $0 到未来 1 年半内变成 $81B 或 $91B）。 这就是为什么我说 $SIVE 看起来像未来的 $75B 体量的 $LITE，会在未来几年内实现。 这一切都将在明年上演。 而它现在的市值，还不到一家拿着 $50M 采购协议、把 Sivers 的激光器买回来重新包装的公司。
- English · 原文@_king142 $SIVE was probably the most recent visible laser chokepoint that’s still being rerated since seem to be in every $NVDA NVLink CPO ecosystem partner + merchant/asic cpo supply chains. Maybe passive optical components supply chains in CPO next.中文 · 翻译@_king142 $SIVE 很可能是最近一个比较显眼的激光瓶颈环节，而且目前仍在被重新估价——因为它似乎在每一条 $NVDA NVLink CPO 生态系统的合作伙伴里都有出现，同时也在商用/ASIC CPO 供应链里。 也许下一个就是 CPO 里的无源光学组件供应链了。
- English · 原文Just some random notes about $AVGO earnings transcript - Revenue target reiterated ($100B+ 2027, pretty sure markets wanted that to be raised this earning, hence the drop) Remember $NVDA Jensen comments about $MRVL $1T company around networking/connectivity/interconnects? - “So as the TPUs continue to accelerate, there’ll be pressure overall on margins. But the connectivity side, the AI networking side of the business has very rich margins” “Demand for … networking is simply insatiable” Also very positive read through as well for the $LITE and the other players. But for TPU margins it goes down at scale, which is understandable. - “they are placing orders in fairly huge demand, which basically gives us a lot more visibility.. runs all the way to 2028 right now” positive read through on overall AI demand since it’s 2026 now… and orders are out in 2028 - The initial order for 1 gigawatt, which includes XPUs and our networking has been received and will start Delivery in the second half of 2027. for our other two customers, we expect shipments to begin late 2026 and accelerate into 2027. $META custom AI program h2 2027 timelines - “Our revenue, our content per gigawatt will increase. you start putting a lot, you start putting embedding CPU cores into the same XPUs and making those chips basically multi die with lots of hvm.” Just for the GW modelers. - “For OpenAI we have delivered silicon and we are on track for production late 2026” OpenAI custom program timeline - “If you ask about 27 or 28 that will continue to grow. We expect in fact 28 to be a substantial growth from what we are forecasting in 27.” More about the demand ramp, go brrr - “Google, that we expect a diversity of sources from them” Mediatek (2454) primary beneficary, maybe $MRVL. Already expected though Google doesn’t sole source so they don’t get bottlenecked. There’s quite a lot of AI demand visibility way until 2028, which is bullish on the AI sector as a whole. Regardless, Broadcom ends the week +0% lol. TLDR: Strongly bullish AI demand, especially networking. Stocks don’t move in a straight line up, but demand curves 2026-> 2027 -> 2028.中文 · 翻译随手记几条 $AVGO 财报电话会议的零散笔记 - 营收目标重申（2027 年 1000 亿美元以上，市场明显是希望这次财报能上调的，所以股价跌了） 还记得 $NVDA 黄仁勋说过，$MRVL 会是一家围绕网络/连接/互联的 1 万亿美元公司吗？ - “随着 TPU 继续加速，整体利润率会有压力。 但连接这一块，AI 网络业务，利润率非常丰厚。” “对……网络的需求简直是无法满足。” 对 $LITE 和其他玩家也是非常正面的连锁反应。不过 TPU 的利润率随着规模扩大而下降，这可以理解。 - “他们下的订单量相当庞大，这基本上给了我们更多的能见度……现在排单一直排到 2028 年。” 考虑到现在是 2026 年，而订单已经排到 2028 年，这对整个 AI 需求是个正面信号。 - 首个 1 吉瓦的订单——包含 XPU 和我们的网络——已经收到，将在 2027 年下半年开始交付。另外两家客户，我们预计出货从 2026 年底开始，并在 2027 年加速。 $META 定制 AI 项目的时间线是 2027 年下半年。 - “我们的收入，每吉瓦的内容会增加。当你开始大量把 CPU 内核嵌入到同样的 XPU 里，让这些芯片基本上变成带大量 HVM 的多裸片结构。” 这是给那些喜欢算吉瓦模型的人看的。 - “对于 OpenAI，我们已经交付了芯片，并将在 2026 年底按期量产。” OpenAI 定制项目时间线。 - “如果你问 27 年或 28 年，那会持续增长。我们实际上预计 28 年相比 27 年的预测会有大幅增长。” 更多关于需求加速的信息，冲冲冲。 - “对于 Google，我们预计他们会有多元化的来源。” 联发科 (2454) 是主要受益方，可能还有 $MRVL。不过这个预期已经在股价里了——Google 不会只依赖单一来源，以免被卡脖子。 AI 需求的能见度一直延伸到 2028 年，这对整个 AI 板块是看多的。不管怎样，博通这周收盘涨跌持平，哈哈。 太长不看版：强烈看多 AI 需求，尤其是网络部分。股票不会直线向上，但需求曲线从 2026 年→2027 年→2028 年一直往上走。
- English · 原文The difference between NASDAQ and EU listing: $POET: $2.4B MC -> Packages Sivers lasers -> One $50m pre-production contract for warrants > $XFAB: $1.7B MC ->SiC/GaN/MEMS/Silicon Photonics Foundry backed with EU CHIPS ACT, US CHIPS ACT PMT -> Below replacement P/B value -> $NVDA, $NOK direct eval of their pre-commercial SiPH foundry, volume ramping 2027/2028 -> $XFAB leading high-volume scaling of Europe's photonic supply chains as the foundry, with IMEC/CEA-Leti, Ligentec, Smart photonics, PHIX Photonics, Luceda Photonics, and Europe's photonic players under it. -> Leading customers like $NVTS, $POWI, Lite-On -> US from Dpt. of Commerce: "the only high-volume SiC foundry in the U.S."中文 · 翻译纳斯达克与欧盟上市的核心区别： $POET：24 亿美元市值 → 封装 Sivers 激光器 → 有一份 5000 万美元的预生产合同（含权证） > $XFAB：17 亿美元市值 → 碳化硅/氮化镓/MEMS/硅光子代工厂，受欧盟芯片法案、美国芯片法案 PMT 支持 → 市净率低于重置成本 → $NVDA、$NOK 直接评估其量产前的硅光子代工厂，2027/2028 年产能爬坡 → $XFAB 作为欧洲光子供应链规模化量产的主导代工厂，旗下有 IMEC/CEA-Leti、Ligentec、Smart Photonics、PHIX Photonics、Luceda Photonics 以及欧洲各大光子厂商 → 主要客户包括 $NVTS、$POWI、Lite-On → 美国商务部认定："美国唯一一家量产级碳化硅代工厂"
- English · 原文@WEB3_furture Great article! I actually didn't expect $NVDA to partner and take a stake in $MRVL this year. But I expected Marvell to be a compelling idea anyway from their upcoming ASIC/connectivity revenue opportunites.中文 · 翻译@WEB3_furture 写得好！我其实没想到今年 $NVDA 会和 $MRVL 合作并入股。 不过我本来就预期 Marvell 会是个有吸引力的标的，出于他们即将到来的 ASIC/连接业务收入机会。
- English · 原文@siflower That's $NVDA directly. I'm talking about the ASIC/merchant NVLink CPO ecosystem partners. Where $SIVE looks like they're the supplier to everyone.中文 · 翻译@siflower 那是 $NVDA 本身。我说的是指 ASIC/商业 NVLink CPO 生态系统里的合作伙伴。 $SIVE 看起来像是所有人的供应商。
- English · 原文its supply chain confirmation, I knew $NVDA was an investor in Ayar, so Id assume they wanted some strategic collaboration like NVlink ecosystem. $AMD also invested in Ayar, so $AMD going with $GFS for CPO also kinda put 1+1 together with $SIVE through Ayar. Mediatek and $INTC turns out to be investors in Ayar, Mediatek does Google ASICs. So if you follow this logic, maybe theres more announcements coming soon with $SIVE in $GOOGL supply chains next.中文 · 翻译供应链已经确认了。我之前就知道 $NVDA 是 Ayar 的投资方，所以我觉得他们应该是想在 NVLink 生态这类方向上寻求战略合作。 $AMD 也投了 Ayar，那 $AMD 找 $GFS 做 CPO 这事儿，再加上 Ayar 这边的关系，基本上把 $SIVE 串起来就很容易推理了。 联发科和 $INTC 原来也是 Ayar 的投资方，联发科还在做谷歌的 ASIC。 所以按这个逻辑推下去，接下来可能很快就会有 $SIVE 进入 $GOOGL 供应链的更多消息。
- English · 原文@OGCapital25 @Chi_w_wong It's expected Celestial and Lightmatter try and multi source. But maybe for gen-1 my guess is a lot sole source / primary source with $SIVE for the $NVDA CPO NVlink ecosystem. Nvidia has their own program with $LITE and $COHR.中文 · 翻译@OGCapital25 @Chi_w_wong 预期 Celestial 和 Lightmatter 会尝试多源供应。 但可能第一代产品，我猜大多是独家或主供，用 $SIVE 来服务 $NVDA 的 CPO NVlink 生态系统。 英伟达自己也跟 $LITE 和 $COHR 有合作项目。
- English · 原文$NVDA CPO ecosystem = $MRVL, Lightmatter, Ayar (joined today) $SIVE 100% confirmed laser supplier to Ayar. $SIVE high confidence laser supplier to Marvell Celestial, Lightmatter. Nvidia, Marvell, Lightmatter, Ayar all use $GFS, which Sivers is reference laser. Sivers = laser supplier to entire NVLink CPO ecosystem. People were doubting Sivers connections to Nvidia, but Ayar joining is clearest tie of $SIVE to $NVDA.中文 · 翻译$NVDA 的 CPO 生态系 = $MRVL、Lightmatter、Ayar（今天刚加入） $SIVE 百分百确认是 Ayar 的激光供应商。 $SIVE 有高置信度是 Marvell Celestial、Lightmatter 的激光供应商。 Nvidia、Marvell、Lightmatter、Ayar 全都用 $GFS，而 Sivers 是它们的参考激光。 Sivers = 整个 NVLink CPO 生态系的激光供应商。 之前有人质疑 Sivers 跟 Nvidia 的关系，但 Ayar 加入就最清楚地说明了 $SIVE 和 $NVDA 之间的关联。
- English · 原文GUESS WHAT ANON? After today’s new news with Ayar joining Nvidia NVLink fusion. $SIVE is now the laser source for likely: The entire Nvidia’s NVLink CPO listed supply chain ecosystem partners. From Marvell Celestial, Lightmatter, and now Ayar today (the three listed in NVLink CPO). This is why I call $SIVE a structural photonics laser chokepoint over CPO and now Nvidia ecosystem supply chains. -> Celestial was likely a direct customer to Sivers, not through Poet. (2023 investor presentation mapping), then bought by Marvell. -> Lightmatter was also listed there as a customer in 2023 investor presentation deck mapping. And… Guess what else? Then they all happen to use GlobalFoundries. Which Sivers is now the GFS silicon photonics foundry-level reference laser (also new news yesterday). Supply chain mapping all starting to make sense now anon? Sivers is also likely now the primary laser source for Ayar after they removed Macom/Lumentum their laser supply chain section (now just gfs/sivers), as a cherry on top. Algorithms completely miss this type of image based mapping. After this announcement, I personally think current valuations are very undervalued: Given Sivers now holds one of the most important structural laser chokepoint over Nvidia CPO NVLink ecosystem supply chains.中文 · 翻译猜怎么着，匿名君？ 今天出了新消息，Ayar 加入了 Nvidia 的 NVLink 融合。 $SIVE 现在很可能是以下领域的激光源： 整个 Nvidia NVLink CPO 列表里的供应链生态系统合作伙伴。 从 Marvell Celestial、Lightmatter，到今天公布的 Ayar（NVLink CPO 里列出的这三家）。 这就是为什么我把 $SIVE 称为 CPO 乃至现在 Nvidia 生态系统供应链上一个结构性的光子学激光瓶颈。 → Celestial 很可能曾是 Sivers 的直接客户，而不是通过 Poet（2023 年投资者演示文档里的映射），后来被 Marvell 收购了。 → Lightmatter 在 2023 年的投资者演示文档映射里也被列为客户。 而且……猜猜还有啥？ 然后它们碰巧都用了 GlobalFoundries。 而 Sivers 现在是 GFS 硅光子学晶圆厂级别的参考激光源（也是昨天的新消息）。 供应链映射开始说得通了，匿名君？ Sivers 现在很可能也是 Ayar 的主要激光源了，因为 Ayar 把 Macom/Lumentum 从它们的激光供应链部分移除了（现在只剩 gfs/sivers），这简直是锦上添花。 算法完全捕捉不到这种基于图像的映射。 在这则公告之后，我个人认为当前的估值被严重低估了： 鉴于 Sivers 现在手握 Nvidia CPO NVLink 生态系统供应链里最重要的结构性激光瓶颈之一。
- English · 原文Wow… new extremely transformative news got released today. Making a certain photonics company: The effective upstream laser chokepoint for $NVDA NVLink fusion CPO ecosystem. With their lasers now in Nvidia’s optical infrastructure supply chains. Can anyone guess the name?中文 · 翻译哇……今天释放了一条极具变革性的新消息。 让某家光子学公司成为了： $NVDA NVLink 融合 CPO 生态系统中有效的上游激光瓶颈。 它们的激光器现在已经进入英伟达的光学基础设施供应链。 有人能猜出名字吗？
- English · 原文@beauty_oe I’m aware! It’s really big news to have $SIVE as the laser supplier to $NVDA nvlink fusion ecosystem. Marvell joined few months ago and look what happened to their market cap after that. I’ll cover it later today!中文 · 翻译@beauty_oe 我知道！$SIVE 成为 $NVDA nvlink 融合生态系统的激光供应商，这真是个重磅消息。 Marvell 几个月前才加入，看看他们之后的市值涨成什么样了。 我今天晚点会详细聊聊这个！
- English · 原文@sdinakar7 I mean a few hundred billion dollar company ran 35% just because of a comment rather than fundamentals. So short term a bit overextended? He did say $1T though so that’s a 4x if you have the patience and believe in Jensenz中文 · 翻译@sdinakar7 我是说，一家几千亿美元的公司就因为一句评论涨了 35%，而不是因为基本面。 所以短期来看有点过度上涨了吧？不过他确实说了 $1T，所以如果你有耐心并且相信 Jensen 的话，那就是 4 倍。
- English · 原文Tbh $XFAB lowkey reminds me of early $TSEM. Just sub <$2B MC. You basically never find a company with $NVDA and $NOK actively validating your pre-commercial silicon photonics foundry… (photonixFAB) While getting CHIPS act/Gov grants to subsidize capex. While leading the Europe’s effort to build a photonics supply chain. Feels like that alone would justify valuations… but you get the power semi SiC/GaN operations for free too and all its assets. CHIPS act 2 is coming out tomorrow, and $XFAB is listed in the photonics blueprints. Did I miss something? Or did markets miss something?中文 · 翻译说实话，$XFAB 的低调劲儿真挺像早期的 $TSEM。 市值连 20 亿都不到。 你基本找不到第二家像这样的公司——$NVDA 和 $NOK 都在积极帮你站台，验证你那还没商用的硅光子晶圆厂（photonixFAB）…… 同时还拿了《芯片法案》/政府补贴来补贴资本支出。 同时还在牵头搞欧洲的光子供应链建设。 感觉光凭这一点就足以撑起估值了……但你还能白捡它的功率半导体 SiC/GaN 业务，以及所有资产。 明天《芯片法案》2.0 就要出来了，而 $XFAB 被列在了光子学蓝图上。 是我漏掉了什么吗？ 还是市场漏掉了什么？
- English · 原文DID YOU LISTEN ANON? Reuters: New Sivers x GFS strategic collaboration. $SIVE has now announced its lasers will be integrated into reference designs built on Globalfoundries Silicon Photonics Platform. For pluggable optical transcivers, CPO, and SiPH. This is fundamentally the most groundbreaking news for Sivers in history. As Broadcom, Nvidia, Marvell, AMD, and anyone who goes through GFS silicon photonics has Sivers embedded as a default laser route. I personally think this news alone should easily 2x or 3x Sivers market cap over the medium term, given how fundamental this is to their revenue. To have Sivers be the standard laser route for the many hyperscalers that use the world's leading photonics foundry.中文 · 翻译兄弟们，你们听到了吗？ 路透社：Sivers 与 Globalfoundries 达成新战略合作。 $SIVE 现在宣布，它的激光器将被集成到基于 Globalfoundries 硅光平台构建的参考设计中。 用于可插拔光收发器、CPO 和硅光。 这对于 Sivers 来说，从根本上讲，是公司历史上最具突破性的消息。 博通、英伟达、迈威尔、AMD，以及任何通过 GFS 硅光平台的公司，都将把 Sivers 的激光器作为默认路线。 我个人认为，单凭这条消息，从中期来看，应该就能轻松让 Sivers 的市值翻倍或三倍，因为这对他们的营收来说是根本性的利好。 让 Sivers 成为众多使用全球领先光子学晶圆厂的超大规模企业的标准激光器路线。
- English · 原文The most consequential event of an entire company’s history. Got released today with a photonics player. Making them the functional standard laser for CPO, Pluggables, and SiPH. For companies like $NVDA, $AVGO, $AMD, to $MRVL using the foundry. Does anyone know the name?中文 · 翻译这可能是整个公司历史上最关键的事件。 今天，一家光子学公司发布了它。 让它成为 CPO、可插拔模块和硅光子的功能性标准激光器。 对于像 $NVDA、$AVGO、$AMD，到 $MRVL 这类使用代工的公司来说。 有人知道名字吗？
- English · 原文@wangxindian @StormDirac I’m not sure people realize the gravity of this news with $GFS yet. It’s probably more fundamentally groundbreaking than $SIVE + $JBL. As $AVGO, $MRVL, $NVDA or anyone who goes through GFS silicon photonics has $SIVE embedded in the reference as default.中文 · 翻译@wangxindian @StormDirac 我感觉大家还没意识到 $GFS 这条新闻的分量。 它可能比 $SIVE + $JBL 更根本性地具有突破性。 因为所有经过 GFS 硅光子的公司，像是 $AVGO、$MRVL、$NVDA，默认的参考设计里都嵌着 $SIVE。
- English · 原文$NVDA Jensen Huang: “ $MRVL the next $1T company ladies and gentlemen “. Marvell is currently trading at $191B. I have positions in Marvell… but how much faith do we have in Jensen for the 5x? https://t.co/II4DTZ5Z9D中文 · 翻译$NVDA 黄仁勋： "各位先生女士，$MRVL 是下一家市值突破 1 万亿美元的公司。" Marvell 当前市值约 1910 亿美元。 我持有 Marvell 的头寸……但我们对黄仁勋这个 5 倍预测的信心有多少呢？ https://t.co/II4DTZ5Z9D
- English · 原文Appreciate the more neutral coverage by Reuters and Bloomberg on $XFAB today. Although it would be nicer to focus more on the structural thesis presented… Around 800 vdc power semis $NVDA exposure + with ongoing $NVDA / $NOK evaluations for photonics. And around CHIPS Act semiconductor sovereignty as the near term catalyst. Rather than around volatility from novel information synthesis. I spend a lot of time looking at regulatory filings to find compelling things market missed you know…中文 · 翻译感谢路透社和彭博社今天对 $XFAB 做了更中立的报道。 虽然如果能更聚焦于所说的结构性逻辑会更好…… 大约 800 伏直流电功率半导体 $NVDA 敞口 + 同时 $NVDA / $NOK 在光电子领域的评估还在进行中。 还有《芯片法案》下的半导体自主权是近期催化剂。 而不是围绕新信息整合带来的波动。 我花很多时间翻监管文件，就是为了找出市场忽略的那些有意思的东西，你懂的……
- English · 原文Bro media… how is $XFAB a meme stock? Can you not repeat the same mistake with $RPI this time? They’re literally getting CHIPS ACT funding from the EU because of how critical they are. And have $NVDA / $NOK evaluating their SiPH side of things, while they traded at a low ~1.28 P/B. This just reminded me of $SOI low p/b but high growth verticals out of legacy segment drag. $XFAB was literally mentioned for CHIPS ACT 2 next week in the blueprints… Which focuses around photonics. The main revenue ramp was around power semis with $NVDA pushing 800 vdc. So $NVTS, $POWI, $WOLF and everyone have been taking off recently. Markets just missed $XFAB, because they’re a lesser known foundry in power semis…But US Dpt. Of commerce pointed them out as the only high volume SiC foundry in the US 2Y ago. I just happened to point out the connections. Just because you don’t understand something, don’t just go call it a “meme stock” with price detached from fundamentals.中文 · 翻译Bro media… $XFAB 怎么就成 meme 股了？ 这次可别再在 $RPI 上犯同样的错误好吗？ 它们可是因为太关键，直接从欧盟拿到了 CHIPS ACT 资金。 而且 $NVDA / $NOK 正在评估它们的 SiPH 业务，同时它们交易价才低到大约 1.28 倍市净率。 这让我想起了 $SOI，市净率低但高增长垂直领域正脱离传统业务板块的拖累。 $XFAB 可是被明确写在下周 CHIPS ACT 2 的蓝皮书里……而那蓝皮书主要聚焦光子学。 主收入增长引擎在于功率半导体，因为 $NVDA 在推 800 伏直流电。 所以 $NVTS、$POWI、$WOLF 和大家都最近起飞了。 市场只是错过了 $XFAB，因为它们在功率半导体里是个不太知名的代工厂……但美国商务部 2 年前就指出，它们是美国唯一高产能的 SiC 代工厂。 我恰好指出了这些关联。 别因为你搞不懂，就给人家扣个 "meme 股" 的帽子，说价格脱离基本面。
- English · 原文@MalteAnkan20 $XFAB is more tethered to power semis growth. With silicon photonics upside optionality depending on how $NVDA / $NOK evaluations go中文 · 翻译@MalteAnkan20 $XFAB 跟功率半导体的增长绑定得更紧。 硅光子这块的上行期权，取决于 $NVDA / $NOK 的评估进展。
- English · 原文$XFAB (photonics + power semis) is an interesting long idea at $1.28B MC, that I took positions in. Given EU CHIPS act 2 is today as the catalyst for European photonics players. > 800 VDC power semi exposure to $NVDA push through $NVTS + $POWI > Silicon Photonics / CPO exposure with $NVDA as evaluation stage for high volume manufacturing (optical transceivers/switches) > The only high-volume SiC foundry in the US. > One of the critical MEMS foundries > ~1.29 P/B, which was around what $SOI was sitting at when I went long. Depressed valuations due to legacy drag > ~6.5-8.5 fwd p/e 2028 personal est. > backstopped by Government: - EU CHIPS act, $128M Euros - US CHIPS act $50M PMT (department of commerce). With likely more coming (just signals critical importance to Western supply chains). So at a certain point with all the grants, they’re just getting the capex funded by the Governments. EU CHIPS act 2 is coming out this week, and I’m gonna go ahead and guess $XFAB might get included given they were before, and this package is specifically targeting photonics. ~$1.3B MC seems compelling to me if it can pull a Soitec reversal (low p/b, very high growth segments, auto legacy drag). As for the $NVDA silicon photonics relationships it’s under “photonixFAB”. Markets probably missed this silicon photonics relationship (like $TSEM when I went long) with Nvidia since XFab leads this… Just under a different name. For power semis, XFAB is named for SiC + $NVTS. In PCN-22181, $POWI explicitly names XFAB as its foundry. Given its exposure to power semis and photonics as growth, low P/B, gov backstop (of course dyor, just sharing my personal thoughts) Thought it personally seemed compelling.中文 · 翻译$XFAB（光子学 + 功率半导体）在市值 $12.8 亿的情况下，是一个值得关注的多头思路，我已经建仓。 因为今天 EU CHIPS 法案 2 出炉，对欧洲光子学玩家是个催化剂。 > 800 VDC 功率半导体业务，通过 $NVDA 传导到 $NVTS + $POWI > 硅光子 / CPO 业务，$NVDA 正在评估其大规模量产潜力（光收发器/交换机） > 是美国唯一一家大规模量产 SiC 的代工厂。 > 也是关键 MEMS 代工厂之一 > 市净率约 1.29 倍，差不多是我当时做多 $SOI 时的水平。受传统业务拖累，估值偏低 > 我个人估算 2028 年前瞻市盈率约 6.5-8.5 倍 > 有政府背书： - EU CHIPS 法案，1.28 亿欧元 - 美国 CHIPS 法案，5000 万美元 PMT（商务部）。 后续可能还有更多（这本身就说明它对西方供应链至关重要）。 所以到某个时间点，算上所有补贴，它们的资本开支基本就是政府买单了。 EU CHIPS 法案 2 这周就要公布，我猜 $XFAB 大概率会被纳入，因为它们之前就在名单里，而且这个方案专门针对光子学。 ~$13 亿的市值，如果它能像 Soitec 那样来个反转（低市净率、极高增长板块、汽车传统业务拖累），我觉得很有吸引力。 至于和 $NVDA 在硅光子方面的合作，项目名称是“photonixFAB”。 市场很可能忽略了 XFab 主导的这层硅光子合作（就像我当时做多 $TSEM 时一样），只是换了个名字存在。 在功率半导体方面，XFAB 就是 SiC 的代名词，和 $NVTS 也是关联的。在 PCN-22181 文件中，$POWI 明确把 XFAB 列为它的代工厂。 考虑到它兼具功率半导体和光子学两大增长点、低市净率、有政府兜底（当然，你自己研究，我只是分享个人看法） 我个人觉得挺有吸引力的。
- English · 原文There’s one very compelling name someone called out. That I ended up taking positions in for power semi exposure. Heavily tied to $NVDA but not directly mentioned like $NVTS. Can anyone guess?中文 · 翻译有人提了个特别有意思的标的。 我后来也建仓了，拿来布局功率半导体方向。 跟 $NVDA 关系很深，但又不像 $NVTS 那样被直接点名的。 有人能猜出来吗？
- English · 原文And now… $MU finally hits a $1 Trillion marketcap. I did say this looks like the next $NVDA given how memory demand looks structural with AI. This stock probably made a lot of millionaires going from $80 to $887. https://t.co/5VFdvcuu2c中文 · 翻译现在……$MU 终于站上了 1 万亿美元市值。 我之前就说过，考虑到 AI 带来的内存需求是结构性的，这看起来就是下一个 $NVDA。 这只股票从 $80 涨到 $887，估计造了不少百万富翁。https://t.co/5VFdvcuu2c
- English · 原文When I do my supply chain mapping... $SIVE is just so critically important to so many frontier industries. I'm not sure people are fully aware yet. $SIVE -> $AAPL, $NOK, $RTX (us defense contractrs), $YSS (golden dome). Not including Ericcson + others. Then you have $SIVE -> $AEVA -> Boston Dynamics / $NVDA self-driving architectural standards. Then you have $SIVE -> Celestial / $POET / Lightmatter / Lightelligence / Ayar / $JBL, and many others for CPO/1.6T. Those go to GUC/ALCHIP/Marvell -> hyperscalers. Then there's a ton more... Like their US Gov CHIPS Act work that are secretive. Normally with stuff like $POET, it's like "hey" you have 1 customer with a $50m purchase order, we know where that's going. Not some Swedish company with a smaller valuation, going into everywhere from Space, Robotics, AI, consumer segments...中文 · 翻译在做供应链梳理的时候…$SIVE 在很多前沿产业里都处于绝对关键的位置。 我感觉大家还没完全意识到这一点。 $SIVE → $AAPL、$NOK、$RTX（美国国防承包商）、$YSS（金穹）。还不包括爱立信和其他。 然后还有 $SIVE → $AEVA → Boston Dynamics / $NVDA 自动驾驶架构标准。 接着是 $SIVE → Celestial / $POET / Lightmatter / Lightelligence / Ayar / $JBL，还有很多其他做 CPO/1.6T 的公司。 这些又通到 GUC/ALCHIP/Marvell → 超大规模云厂商。 还有更多…比如他们跟美国政府《芯片法案》相关的保密项目。 通常像 $POET 这种，你会说"嘿"，它有一个客户下了 $5000 万的采购订单，我们知道钱会流向哪里。 但一家瑞典公司，估值更小，却遍布太空、机器人、AI、消费电子等各个领域。
- English · 原文Fun fact: Lot of the same companies are often used across different supply chains. One likely example is: $SIVE as the upstream laser supplier to Boston Dynamics via: Sivers -> $AEVA FMCW (CW DFB lasers) -> LG Innotek -> Boston Dynamics. I actually personally liked Aeva for 4D AI first. Just so happened to find out Sivers was their high confidence laser supplier for 4D FMCW lidar. So you actually get robotics exposure with photonics while the same CW lasers used for hyperscaler AI DCs. Near term revenue ramp though it's probably $SIVE supplying laser volume ramp for $NVDA self-driving car related architectures though Aeva. Humanoids are probably later in 2028? You can always get more indirect exposure like MU with memory or $INTC with edge CPUs, but of course there's more direct exposure out there. Think I've already covered a lot of names in the past like $VPG or Harmonic Drive. But hilariously enough CPO players like $SIVE are a core part of frontier physical AI development.中文 · 翻译有意思的是，很多公司其实会同时出现在不同的供应链里。 一个可能的例子是：$SIVE 作为波士顿动力的上游激光供应商，路径是这样的： Sivers → $AEVA FMCW（连续波 DFB 激光器）→ LG Innotek → 波士顿动力。 坦白说，我一开始是因为 4D AI 才看好 Aeva 的。 后来才发现 Sivers 是他们 4D FMCW 激光雷达的高确信度激光供应商。 所以你看，你可以通过光子学同时参与机器人赛道，因为同样的连续波激光器也用于超大规模 AI 数据中心。 不过要说到近期的收入爬坡，大概率是 $SIVE 通过 Aeva 为 $NVDA 自动驾驶相关架构提供激光器的量产规模。 人形机器人可能要到 2028 年之后？ 你也可以找一些更间接的参与方式，比如 MU 做内存、$INTC 做边缘 CPU，不过当然市场上也有更直接的标的。 我记得我之前已经聊过不少名字了，比如 $VPG 或者 Harmonic Drive。但搞笑的是，像 $SIVE 这样的 CPO 玩家，反而是前沿物理 AI 开发的核心部分。
- English · 原文@Fletche33881557 Nope, $NVDA said $3-4T annual spend. Not value. https://t.co/sKg5gA2RZO中文 · 翻译@Fletche33881557 不，$NVDA 说的是每年 3-4 万亿的支出。不是价值。https://t.co/sKg5gA2RZO
- English · 原文AI capex spend is expected to go to "$3 to $4 trillion annually" by 2030 from $NVDA Jensen Huang projections. You're not bullish enough. And it might be a good idea to stay exposed + own the keys of the AI Kingdom: -> $AXTI controls the materials buildout with photonics. -> $SOI controls the AI buildout with silicon photonics. -> $SIVE controls laser chokepoints for CPO. -> $IQE controls Western epiwafer supply chains for photonics. All these started off as tiny companies, yet the trillions of projected capex gradually upward to them. There's many more in other industries as well. -> AI Capex flows to Neoclouds like $NBIS. -> AI Capex flows to memory like $MU and $SNDK. And many of the "commodity" materials or "science projects" for the past 20 years now a sudden shift in exponential TAM expansion. We're witnessing the next industrial revolution with Artificial Intelligence + Physical AI.中文 · 翻译根据 $NVDA 黄仁勋的预测，到 2030 年，AI 资本支出预计将达到“每年 3 到 4 万亿美元”。 你还是不够乐观。 保持敞口、掌握 AI 王国的钥匙或许是个好主意： -> $AXTI 通过光子学掌控材料建设。 -> $SOI 通过硅光子学掌控 AI 建设。 -> $SIVE 掌控 CPO 的激光瓶颈环节。 -> $IQE 掌控西方光子学外延片供应链。 这些公司最初都很小，但预计数万亿美元的资本支出会逐渐流向它们。 其他行业还有更多例子。 -> AI 资本支出流向像 $NBIS 这样的 Neocloud。 -> AI 资本支出流向像 $MU 和 $SNDK 这样的存储器。 而且，很多过去 20 年里被视为“大宗商品”材料或“科研项目”的东西，突然间迎来了指数级的 TAM 扩张。 我们正在见证人工智能+物理 AI 带来的下一次工业革命。
- English · 原文All right chat, crowdsourcing your #1 highest conviction (10x only) stock long for the Power Semi trade. Especially given $NVDA pushing shift to 800 VDC. Stuff like $NVTS or $WOLF, but high-beta, 10x potential only. Anywhere around the world. What's your pick? https://t.co/Y10KR6HLVC中文 · 翻译好了大伙儿，来集思广益一下——你心目中 Power Semi 板块里最有信心（只要 10 倍）的做多标的。 特别是考虑到 $NVDA 在推动转向 800 VDC。 像是 $NVTS 或 $WOLF 这种，但要高 Beta、只求 10 倍潜力。全球范围都行。 你选哪个？https://t.co/Y10KR6HLVC
- English · 原文@_juicebox99_ I don't see FOCI becoming a $50B+ company. They focus on one thing like optical components + FAU for $TSM / $NVDA, then do it at scale. So that would likely drives structural re-rating related to MC. Not quite the same as the US model of TAM expanding into $100B+ companies.中文 · 翻译@_juicebox99_ 我不觉得 FOCI 能成为市值超过 500 亿美元的公司。 他们专注的就是单一赛道，比如为 $TSM / $NVDA 提供光学元件和 FAU，然后规模化生产。所以这很可能带来与市值相关的结构性重估。 这和美国的 TAM 扩张到千亿美元级别的公司模式，不太一样。
在其他光模块产业链标的上也有持续观点
NVDA 之外，以下标的同样收录于其观点库：