全部推文
共 6775 条 · 第 160 / 226 页 · 滚到底自动加载更多
-
质疑特斯拉Optimus量产进度，认为最新言论暗示延迟。
@jayjay12246 我的意思是，鉴于埃隆两天前表示他们已在弗里蒙特工厂正式开始了“Optimus Gen 3 (V3) 人形机器人的大规模生产(Mass Production)”，我以为会早一些。但他今天的言论暗示存在延迟，所以真的不太确定。 https://t.co/qxSZxmYsbt
英文原文
@jayjay12246 I mean I thought it would be earlier given Elon said 2 days ago that they've officially started mass production "Mass Production Optimus Gen 3 (V3) humanoid robot at its Fremont". But his statements today imply delays, so really not quite sure. https://t.co/qxSZxmYsbt
-
推荐欣兴电子，因其布局广泛且鲜为人知。
@genZinvest0r 说得在理！我只是想把欣兴电子(Unimicron) $UMC 放到大家的视野中，因为绝大多数人以前都没听说过这只股票。在我看来，这似乎是一个极佳的做多(long)机会，因为它在位置布局上涉及众多不同的建设(buildouts)项目。
英文原文
@genZinvest0r That’s fair! I’m just putting Unimicron on people’s radar since this a stock the vast majority of people have not heard about before. It just seems like an incredible long to me since positionally it sits in so many different buildouts
-
推荐IBKR交易全球市场，TTMI作为替代，并提示市场机会众多。
我在美国，但经常四处旅行。$IBKR 可以访问大多数市场。话虽如此，如果你想避免外汇费用，$TTMI（一年涨幅267%）等美国标的可以提供类似的敞口，但显然并不完全相同。此外，我讨论的话题很多，市场上还有大量其他机会。
英文原文
US, but I travel around a lot. $IBKR has access to most markets. That being said there are similar US plays like $TTMI (up 267% 1Y) for similar exposure if you want to avoid foreign fees but obviously not quite the same. Also I talk about a lot of things, there's tons of other opportunities out there
-
NBIS长期增长强劲但短期受增发压制，Q3或迎拐点。
关于 $NBIS，这是一只长期表现惊人的股票，因为其旗下四家公司的业务同比增长均超过100%，其中主营业务更是增长了700%+。 然而短期内，他们拥有超过20亿美元的活跃 ATM（增发融资），这带来了巨大的抛压。 我预计第三季度情况会好转，届时市场开始提前交易其70亿美元 ARR（年度经常性收入）目标。Clickhouse 150亿美元的估值非常看涨，此外还应关注 $UBER 与 avirde 合作的 Robotaxi（自动驾驶出租车）进展。除此之外，近期进展相当缓慢，缺乏引人注目的亮点。
英文原文
As for $NBIS, amazing long term stock since all 4 of their companies are growing 100%+ y/y with their main business 700%+. However near term, they have an active $2B+ ATM, which introduces tons of selling pressure. I’d expect things to pick up q3 when markets start frontrunning their $7B ARR target. Clickhouse $15B valuation was incredibly bullish + should pay attention to $UBER Robotaxi ramp with avirde. That aside been pretty slow without much flashy developments.
-
欣兴电子处于关键供应链瓶颈，但西方投资者难以交易。
@PhotonCap 欣兴电子(Unimicron)身处众多未来建设计划及潜在瓶颈环节之中，因此对我来说也极具吸引力。话虽如此，由于它在台湾证券交易所(TWSE)上市，西方投资者很难获得交易权限。
英文原文
@PhotonCap Unimicron sits in so many future buildouts and potential bottlenecks so it’s incredibly interesting to me as well. That being said it’s hard for Westerners to get access to since it trades on TWSE
-
AEHR具登月潜力，Sonoma订单或关联MU HBM4，但尚无确凿证据。
@Lotharinvest 我了解 $AEHR，正如你所说，它具有“登月式”（moonshot）的结构（例如，最近 550 万美元的 Sonoma 订单与 $MU HBM4 16-high 资格认证阶段相符）。目前尚无确凿的关联证据。
英文原文
@Lotharinvest I’m aware of $AEHR, it has the structure of a moonshot (eg. $5.5M Sonoma order recently in line with $MU HBM4 16-hi qualification phase) as you mentioned. No confirmed links yet.
-
英特尔是政策驱动的非对称多头，短期期权风险不影响长期逻辑。
我不认为有人预期 $INTC 能在1-2个季度内交付所有成果。这是最不对称的多头头寸之一，因为这是一场等待游戏，且有美国政府为建立其晶圆代工厂提供的无限后盾。我会说这是对政策的多头押注，而非基本面。就时机而言，大量短期深度虚值看涨期权（OTM Calls）的情况有点令人担忧。但如果你做多未来几年，这其实并不重要。
英文原文
I don't think anyone expected $INTC to deliver everything in 1-2 quarter. It's one of the most asymmetrical longs there are, since it's a waiting game with infinite backstop from US Gov to buildout their foundry. I'd say it's a long on policy, not fundamentals. In terms of timing the amount of short term way OTM calls are a bit worrysome. But it shouldn't really matter if you're long for the next few years.
-
看好欣兴电子，作为AI供应链关键瓶颈，受益于存储及先进封装需求。
我持有市场忽视的$MU相关头寸： 欣兴电子(Unimicron)。 这家市值160亿美元的公司是极佳的长线标的，因为它是以下领域的基础供应商： - $MU和三星的存储基板 - $NVDA的Rubin/Blackwell架构 - 台积电(TSMC)的CoWoS工艺 - 苹果（iPhone和M系列芯片） - 谷歌TPU - AMZN/超大规模云厂商的ASIC - 英特尔的ABF基板 - 购买欣兴基板的Amkor - 以及全球所有半导体。 欣兴电子是全球AI供应链中最显著的“瓶颈”之一，因为它提供面向2026-2028年的未来核心部件： - 用于HBM4/存储（DRAM/NAND）+ GPU/CPU的IC基板 - 台积电CoWoS中的“S”（基板） - 用于CoPoS的玻璃核心基板 以及当前iPhone等设备所需的所有基础部件。 它处于存储瓶颈、CoWoS瓶颈以及超大规模云厂商ASIC互连瓶颈的中心。 同时也涵盖未来周期的玻璃核心基板等。 AI围绕另一家Micron（欣兴）展开，你可以借此暴露于当今及未来最热门的瓶颈环节。
英文原文
I've hold positions in the $MU that markets sleep on: Unimicron. This one is an incredible long at $16B as they're foundational for: - $MU and Samsung memory substrates - $NVDA Rubin/Blackwell - TSMC for CoWoS process - Apple (iPhones and M-series) - Google TPUs - AMZN/Hyperscaler ASICs - Intel for ABF substrates - Amkor that buys Unimicron substrates - and all the world's semis. Unimicron is one of the most significant "choke points" in the global AI supply chain as they provide the future: - IC Substrates for HBM4 / Memory (DRAM/NAND) + GPUs/CPUs - The "S" (substrate) in TSM CoWoS - Glass Core Substrates for CoPoS for 2026-2028 and all the foundational parts for customer devices like IPhones and semiconductors used today. They're in the center of the memory bottleneck, CoWoS bottleneck, and interconnects bottleneck for Hyperscaler ASiCs. As well as the future cycles with Glass Core substrates and more. AI revolves around the other Micron and you get exposure to most popular bottlenecks like HBM today and tomorrow.
-
中国AI企业受推理算力限制，凸显英伟达在供应链中的绝对主导地位。
有很多报道称中国科技公司（字节跳动、$BABA 等）在推理算力上正面临困境。 中国可能仍未改变其国内自主可控的目标，但推理限制的严峻现实可能迫使他们做出妥协。 这恰恰证明了 $NVDA 的主导地位，即 AI 企业若没有它便无法运转。
英文原文
There were many reports Chinese tech (bytedance, $BABA, etc) were struggling with inference capacity. China still hasn’t likely changed their domestic self-reliance goal but reality of inference limits likely forced their hand. Just goes to show how dominant $NVDA is that AI firms can’t function without it
-
英特尔股价回落，CFO财报会表现糟糕。
@acchcch $INTC 股价已回到周一的水平。抛开价格不谈，首席财务官(CFO)在财报电话会议上的表现糟糕透顶。
英文原文
@acchcch $INTC price is back to where it was… Monday. Price aside, the CFO sounded horrendous on the earnings call
-
提醒通过东京交易所交易3110.T，券商接入有限且OTC价差大。
@joaofrios 通过东京证券交易所交易 $3110.T。不过，并非每家券商都能接入日本市场。虽然有场外交易(OTC)版本，但我确信其买卖价差会非常大。
英文原文
@joaofrios $3110.T through Tokyo Stock exchange. Not every broker has access to Japanese markets tough. There is an OTC version but I’m sure the spreads are extremely large.
-
日企Nitto Boseki垄断AI关键材料T-glass，巨头争抢供应至2026年。
Nitto Boseki - $NBCLF | 3110: AI极端瓶颈的故事： > 1923年开始纺纱/玻璃纤维 > 做到极致 > 为网球拍创造“T-Glass”（T-玻璃） > 等待100年，所有人都需要它 > $TSM 用于 CoWoS（芯片级封装） > $APPL 用于 M-Series Ultra/Max > $NVDA 用于 Blackwell + Vera Rubin > $TSLA 用于 FSD 汽车内的芯片 > $AMD 用于 MI300/MI400 AI 加速器 > $ANET 用于 800G 到 1.6T 和 3.2T > 苹果、英伟达和台积电争夺供应 > 已售罄至 2026 年 > 苹果、英伟达和高通派团队驻厂，字面意义上向这家纱线制造商乞求玻璃供应 > 目前在半导体用高端 T-glass 市场占据 80-90% 份额 这家价值 42.3 亿美元的垄断企业，全球科技公司都在争夺： 一家随机的日本纱线制造商。 披露：我持有头寸。
英文原文
Nitto Boseki - $NBCLF | 3110: The story of an extreme bottleneck for AI: > Spin Yarn/Glass in 1923 > Be good at it > Create "T-Glass" for tennis rackets > Wait 100Y, everyone needs it > $TSM for CoWoS > $APPL for M-Series Ultra/Max > $NVDA for Blackwell + Vera Rubin > $TSLA for chips inside FSD cars > $AMD for MI300/MI400 AI accelerators > $ANET for 800G to 1.6T and 3.2T > Have Apple to Nvidia to TSM fighting over supply > Sold out through 2026 > Apple, Nvidia, and Qualcomm sends teams live on site to literally beg this Yarn Maker for glass supply > Now is 80-90% market share of the high-end T-glass used in semiconductors The $4.23 Billion monopoly that every tech company in the world is fighting over: A random Japanese Yarn Spinner. Disclosure: I own positions.
-
反思未在META跌至603美元时加大买入力度。
@pepemoonboy 我本应在 $META 跌至 603 美元时更多地通过成本平均法(Cost Averaging)买入。 https://t.co/6LEijFygpy
英文原文
@pepemoonboy I should have cost averaged $META more on the drop to $603. https://t.co/6LEijFygpy
-
POET财务强劲且无债，下跌属过度反应，现价具吸引力。
如果看今天的财报，$POET 处于极佳的位置。当然，这是以牺牲前散户股东为代价的。无债务，约4.35亿美元现金对应10亿美元市值。他们可以利用这笔资金储备为 Celestial 扩大生产/库存。其稀释效应不像 $BMNR 那样，为了向 Mr. Beast 的巧克力棒公司投入2亿美元。所以我认为这次下跌反应过度了，能以低于承销商价格4%以上的价格买入很不错。
英文原文
$POET is in an extremely good spot if you look at the financials today. Of course this is at the expense of former retail shareholders. No debt, ~435M cash to $1B MC. And they can use that war chest to ramp up production/inventory for Celestial. Dilution isn't like $BMNR to put $200m into Mr. Beast's chocolate bar company. So I'd say the drop was quite the overreaction and it's nice to get in 4%+ underneath underwriter prices.
-
POET 暴跌后作者认为估值合理，虽认为时机稍早但选择逢低买入。
$POET 在直接发行 1.5 亿美元股份（发行价 7.25 美元）后暴跌 17%。 考虑到新增 2060 万股，其市值约为 10.2 亿美元，目前手握约 4.2 亿至 4.5 亿美元的现金储备（war chest）。 在我看来，这是一个不错的机会，因为机构发行价为 7.25 美元。 Poet 的光学中介层（optical interposers）是解决磷化铟（InP）瓶颈的真正方案，且拥有超大规模云服务商（hyperscaler）的后门渠道。特别是因为他们向 $MRVL 的 Celestial 光子网络（photonic fabric）供应 Starlight 产品。 我认为公司提前了 5-8 个月，所以我个人没有做多。 但我个人在这个价位买入，以利用这次下跌。
英文原文
$POET has crashed 17% after a direct offering of $150m in shares at $7.25. They’re now sitting on a ~$420-$450m war chest with a ~$1.02B MC factoring in 20.6M new shares. IMO this is a decent opportunity as the institutional offering price was $7.25. Poet’s optical interposers are a genuine bottleneck fix to InP and they have a hyperscaler backdoor. Especially because they supply Starlight to $MRVL Celestial photonic fabric. IMO company is 5-8 months a bit early so I’m not personally long. But I personally ended up buying at this level to take advantage of the drop.
-
AXTI光子学原料受出口禁令影响，数月后或现瓶颈。
@0x_ysn 这与存储芯片关系不大，因为 $AXTI 涉及的是光子学(Photonics)输入端，几个月内会出现瓶颈。我们将看到西方公司在两周前的出口禁令后，还有多少个月的原料/前驱体库存可供消耗。
英文原文
@0x_ysn Not really related to memory since $AXTI is for photonics inp bottleneck in a few months time. We’ll see how many months of feedstock/precursors Western companies have + use up post export-ban 2 weeks ago.
-
博主回应粉丝，承认遗漏了NVMI和CAMT两只股票。
@Mamocap 谢谢，我错过了 $NVMI。有人提到了 $CAMT，我上面那份名单里不小心漏掉了它。
英文原文
@Mamocap Thanks missed out on $NVMI. Someone else mentioned $CAMT, I left it out of the list above by accident.
-
确认CAMT为HBM4明确受益标的。
@markabelsantos 是的，$CAMT 是高带宽内存第四代(HBM4)的明确受益者。
英文原文
@markabelsantos Yeah $CAMT is a clear beneficiary for HBM4
-
澄清MRAM不属于英伟达或美光HBM4堆栈。
@Lotharinvest $MRAM 并不位于 $NVDA / $MU 的高带宽内存第四代(HBM4) 堆栈内部。
英文原文
@Lotharinvest $MRAM does not sit inside the $NVDA / $MU HBM4 stack.
-
Rambus凭HBM4 IP高毛利躺赚三星等大厂，是潜力黑马。
是的，$RMBS 是一匹黑马。三星曾公开表示“Rambus HBM4 IP 解决方案的可用性至关重要”。 他们正趁着存储周期，从三星、SK 海力士和 $MU 身上获利，毛利率高达 81% 以上。 他们虽没有三大巨头那样的定价权，但利润应该只会源源不断。
英文原文
Yeah, $RMBS is a sleeper pick. Samsung went out and said "availability of Rambus HBM4 IP solutions will be critical". They're just profiting off Samsung, Sk Hynix, and $MU off during memory cycle with 81%+ gross profit margins. They don't quite get the same pricing power like the big three do but it should just print.
-
梳理受益于美光及三星的HBM4供应链中小市值美股标的。
存储超级周期供应链要点： 受益于 $MU 和 SK/三星的美国冷门 HBM4/存储阿尔法标的： $VECO ($20亿) - 激光退火 $PLAB ($21亿) - 光刻掩模版 $ADEA ($21.8亿) - 混合键合 $ACLS ($29亿) - 离子注入 $FORM ($64.3亿) - HBM晶圆测试 $ONTO ($106亿) - 封装量测 $AMKR ($131亿) - 先进封装 $RMBS ($135.5亿) - 存储IP $MKSI ($149亿) - 激光子系统 未包含 $TER 和 $KLAC 等巨头，只想聚焦于知名度较低的标的。 还有一些间接受益者如 $SMTC 和 $MTSI。市场肯定尚未完全定价其中部分标的。 如果我漏掉了任何标的请告诉我。 只是想发布这些，以防你们对其中任何一只感兴趣。
英文原文
Memory Supercycle Supply Chain TLDR: US Sleeper HBM4/Memory alpha picks that benefit from $MU and SK/Samsung: $VECO ($2B) -Laser annealing $PLAB ($2.1B) - Lithography photomask $ADEA ($2.18B) - Hybrid bonding $ACLS ($2.9B) - Ion implantation $FORM ($6.43B) - HBM wafer testing $ONTO ($10.6B) - Packaging metrology $AMKR ($13.1B) -Advanced packaging $RMBS ($13.55B) - Memory IP $MKSI ($14.9B) - Laser subsystems Didn't include big players like $TER and $KLAC since just wanted to focus on lesser known ones. There's some indirect beneficiaries too like $SMTC and $MTSI. Markets definitely haven't priced some of these in yet. Let me know if I missed any. Just wanted to publish these in case you find any of them interesting.
-
看好$RMBS，类比$ARM，高毛利且预期估值提升。
$RMBS 是个不错的选择，我持有该头寸。 这让我想起 $ARM，不过是针对高带宽内存(HBM)领域的，在散户眼中几乎无人问津。 它没有像 $MU 那样的定价权，但其向 SK 海力士、三星和美光授权的技术应该会带来丰厚回报。毛利率高得惊人。 我预计其估值倍数(Re-rating)在今年也会提升。
英文原文
$RMBS is a good shout, I have positions in it. Reminds me of $ARM but for HBM, pretty nonexistent in retail eyes. It doesn’t get the same pricing power like $MU but all its licenses to sk Hynix, Samsung, and micron should pay dividends. Gross margins are crazy high Id expect it to get rerated as well over this year
-
对比沃尔玛与美光估值，强调基本面优于价格。
$WMT 的前向市盈率(Forward P/E)为 40.43 倍。 没错，就是沃尔玛，那个浴巾转售商，其营收同比增长 4.9%，仅与通胀持平。 $MU 的前向市盈率为 11.6 倍，营收同比增长 133%，每股收益(EPS)同比增长 319%。 美光(Micron)还有很长的路要走，仅看价格并非最明智的投资方式，因为基本面更为重要。 股票每天都在创历史新高(ATH)。
英文原文
$WMT trades at 40.43 forward P/E. Yes, Walmart the bath towel reseller that's growing revenue 4.9% Y/Y in line with inflation. $MU trades at 11.6 forward P/E growing revenue 133% Y/Y, and EPS 319% Y/Y. There's a long way to go for Micron, looking at prices alone is a not the brightest way to invest since fundamentals are more important. Stocks make ATHs every day.
-
MU因国家安全护城河及低估值，是押注美国政策的最优多头。
人们很容易因认证风险或SK海力士产能爬坡及竞争而过度思考。 $MU拥有全球最深的护城河： 美国国家安全+“美国制造”供应链。 白宫+美国超大规模云服务商不会让$INTC或$MU失败。且其远期市盈率仅为11倍。 这是押注美国政策最轻松的多头机会。
英文原文
It's extremely easy to overthink things with qualification risk or Sk Hynix ramp + competition. $MU has the biggest moat in the world: US national security + Made in America supply chains. The White House + US hyperscalers won't let $INTC or $MU fail. And it's trading at 11 forward p/e. This is the easiest long on US policy.
-
类比 Nvidia 历史，看好美光在 AI 存储超级周期中的重估潜力。
$MU 看起来是下一个 Nvidia。当 Nvidia 市值为 4000 亿美元（现超 4.5 万亿美元）时，市场认为 GPU 只是短期周期。如今的存储芯片也是如此。AI 打破了这一周期。凭借与 $INTC 相同的“美国制造”和白宫支持：不要对美光(Micron)想太多。
英文原文
$MU looks like the next Nvidia. When Nvidia was $400B (now $4.5T+), markets thought GPUs were a short-term cycle. Same with memory today. AI has broken that cycle. With the same "Made in America" and White House backing like $INTC: Don't overthink things with Micron. https://t.co/LBhbVjoKGz
-
分析美光因本土优势及需求结构性转变的看多逻辑。
这只是一个针对 $MU 的极度简化的看多情景。当然也存在看空因素，例如 @MarkosAAIG 提到的 $NVDA Rubin 在 HBM4 认证方面的执行风险。$INTC 的逻辑与此类似，美光位于美国。美国政府可能对海力士/三星征收关税以推动本土使用，加上超 60 亿美元的赠款、数百亿美元的税收抵免，以及超大规模云厂商对本土供应链的需求，即使出现延迟，美光也拥有巨大优势。在需求端，他们 2026 年及 2027 年的订单已售罄。我认为这是一次结构性需求转变，而非周期性波动，尽管人们可以有不同的解读。
英文原文
So this is just an oversimplified bull case scenario for $MU. There's obviously bear cases too, one being HBM4 qualification execution risk for stuff like $NVDA Rubin that @MarkosAAIG mentioned. Similar thesis with $INTC here, Micron is in the US. US government likely tarrifs on hynix/samsung to push for domestic usage, $6B+ in grants, and tens of billions in tax credits, and hyperscalers wanting a domestic supply chain gives Micron a pretty big edge even if there's delays. As with the demand side, they're sold out for 2026 and into 2027. I see a structural demand shift, not a cycle this time around, people can interpret it differently though.
-
TSM前董事长买入MU被视为看涨信号
@GVDInvestor 完全同意，上周$TSM 前董事长在公开市场买入 800 万美元的 $MU 是一个看涨信号。
英文原文
@GVDInvestor Completely agree, having the former chairman of $TSM buy $8m of $MU on the open market last week is a bullish tell.
-
类比英伟达，认为美光科技第二波行情上涨预期强烈。
@Thomas_james_1 这让我想起了英伟达($NVDA)在市值攀升至4万亿美元以上时的情况，当时大家都觉得它会上涨。现在，美光科技($MU)在第二波行情中开始让人感觉如出一辙。
英文原文
@Thomas_james_1 Just reminds me of $NVDA during the rise to $4T+ valuation where everyone kinda knew it would go up. Now it’s starting to feel the same with $MU in round 2.
-
不建议对美光进行择时交易，认为其估值仍有重估空间。
@davesetton 正如其他人所说，我个人不想对 $MU 进行择时交易。最终决定权在你，但在我看来，鉴于其远期市盈率(fwd p/e)仍为11倍，仍有大量的重估空间。
英文原文
@davesetton Like the others said i personally wouldn’t want to time the market with $MU. It’s your decision in the end but in my opinion there’s a lot of re-rating to do given its fwd p/e is still 11.
-
美光估值偏低，机构逢低买入，目标价$672，2027年或达$1.5T市值。
$MU 远期市盈率仍约为 11.6 倍。 它正在预示： - 营收同比增长 133% - 每股收益(EPS)同比增长 319% 这就是所谓的存储超级周期(memory supercycle)的原因。 即使本月已上涨 40%+，它甚至尚未完全完成估值重估(re-rating)。 按当前价格，它仍被估值为一家缓慢增长的大宗商品公司，目标价 $387。 这就是为什么在估值重估过程中，每一次回调都被机构买入，因为高带宽内存(HBM)现在极具价值。 在我看来，鉴于当前的增长/利润率，美光(Micron)至少值得 20 倍远期市盈率，市值约 7500 亿美元，即每股 $672。 到 2027 年，如果美光交易在 1.5 万亿美元市值，我不会感到惊讶。
英文原文
$MU forward P/E is still ~11.6 It's projecting: - 133% Y/Y Revenue Increase - 319% Y/Y EPS Increase This is called the memory supercycle for a reason. Even after a 40%+ this month, it's still hasn't even been fully re-rated yet. At current prices, it would still be valued like a slow growth commodity at $387. This is why every single dip gets bought by institutions during the re-rating process, since HBM is now extremely valuable IMO Micron deserves at least a 20 fwd P/E at today given growth/profit margins, ~$750B so $672/share. In 2027, wouldn't be surprised if Micron was trading at $1.5T.