
EON
1.6K posts






今天刷Serenity的主页,发现她关注了我,惊讶又开心。如果你这一年关注美股半导体,尤其是 AI 产业链、光模块、硅光、存储、CPO、衬底等这些方向,大概率会刷到过她。 我最开始涨粉 1000 多的一篇帖子就来自于引用她关于投资路径的观点。 那篇内容对我来说挺重要的。因为在那之前,我看半导体更多还是看股票代码、看财报、看涨跌。看到她的分析之后,我第一次真正要沿着产业链往下拆,去看每一轮需求爆发时,真正卡住供给的环节在哪里;去看资金轮动方向。 这套思路后来对我影响很大。我选择重仓的是存储,尤其是闪迪、三星、海力士这一条线。事后看,这个方向确实给了我很大的收益。但同时我也很清楚自己其实错过了很多其他环节的多倍增长。 比如光模块、硅光、CPO、部分材料和小市值供应链公司,有些涨幅并不比存储差,甚至更夸张。 非常感谢Serenity帮助我打开了投资视角。她关注的很多方向,一开始都不是市场最拥挤的主线,但后来会随着 AI CapEx 扩散,被资金重新定价。她致力于把一个还没有被充分定价的逻辑提前讲清楚,并且接受市场后续验证。 我关注 Serenity 的时候,记得她好像还只有几万粉丝。现在已经 30 多万了。她的成长本身也很有代表性,在一个信息越来越噪音化的平台上,真正持续输出高密度产业链研究的人,依然会被市场看见。 她最近关注我,对我来说更像一个小小的闭环。 最开始我只是转达她的观点,借着那篇帖子涨了 1000 多粉,也真正开始研究半导体产业链。之后也引用过多篇内容,几个月后,我自己也在这条路上慢慢积累了一点判断,靠存储这条线赚到了一些钱,也经历了很多次波动、卖飞、回撤和复盘。 投资账号的成长,和投资本身很像。 一开始你只是看见别人怎么分析,后来你慢慢形成自己的框架,再后来,你开始用自己的语言把复杂的东西讲给更多人听。 中间一定会错过,一定会犯错,也一定会有很多如果当时的遗憾。但只要你的认知在往前走,这些错过就不是白白错过。它们会变成下一次判断的底层素材。 最后再次感谢 Serenity @aleabitoreddit 。 感谢长期免费分享高密度的产业链研究,也感谢对投资博主价值的这种定义和坚持:不靠晒账户金额制造权威感,而是靠公开 thesis、持续验证和市场结果来证明研究的有效性。 我也会继续关注和跟踪她的研究,继续学习这种从信息发现、产业链拆解到市场验证的能力。希望自己以后不仅能抓住更多机会,也能把复杂的投资逻辑讲得更清楚更有价值。




$TRT - the market just handed me a gift. I'm doubling down on this name. This is an $AEHR in the making and you are selling? I just rotated a +100% swing position from $TE into $TRT on today's dip at $11.78. Morgan Stanley estimates the Powered Shell at 6-14% of every hyperscaler build. That is $2-6bn per GW at stake. Every GaN and SiC device shipped into that layer needs to be tested before it goes into a rack. Both $AEHR and $TRT are major beneficiaries playing that opportunity differently. $AEHR operates at the wafer level. Before the chip is even cut from the wafer, Aehr burns it in. This is the earliest and most specialized test step in the entire production flow. If a GaN or SiC wafer fails here, it never becomes a device. $TRT operates at the back end. After the chip is packaged, Trio-Tech runs burn-in, environmental stress, and reliability validation. This is the final quality gate before a device ships to a customer. Same supply chain. Different layers. Neither replaces the other. The key difference: $AEHR is a US-listed Nasdaq name with wafer-level IP that is difficult to replicate. Higher margin, higher multiple, more institutional coverage. Power semi exposure has become one driver among several as AI processor burn-in takes over. $TRT is the Asian back-end pure-play catching the same wave at a much earlier stage of market recognition. Had a stellar ER with accelerating growth Q1: 58%. Q2 +82%. Q3: +124% revenue growth. $7.8M in orders already confirmed. Revenue run rate approaching $65M annualized. Balance sheet clean. Debt-to-equity O.11. New facility coming online in Malaysia, which is a hot spot. Higher execution risk but also much more room to re-rate if the AI and power semi ramp continues. The Morgan Stanley model shows $2-6bn per GW flowing into the Powered Shell. Every chip in that layer needs both test steps. $AEHR captures the front. $TRT captures the back. The opportunity is large enough for both to win. The difference is that $AEHR sits at a $2.5bn market cap and $TRT sits at $200mn. I am not trading this one for a quick flip. I am holding for an easy x3-x5 within a year. If it keeps dipping, I will buy more. I own both. I am loading up on $TRT. This is still early for $TRT but not for long. I'm long $TRT. NFA, you do you.





Okay chat. I'm curious what your thoughts are on the next 1350%+ 1Y return $SNDK or 2918% 1Y return $AXTI. If you have to choose 1 very high conviction, hyperbolic growth stock... What would it be and why?















I guess, post earnings when $ARM touched $268... $ARM is now #18 on the individual stock list that I went long on that hit 100%-1000%+ YTD? I've lost count TBH. Some others like $LPK and $SIMO and $HPS.A are getting really close now. But feels like I'm one of the few ones out there on X with actual receipts of all the returns + original thesis post.

IT'S REAL! IBKR now has Taiwanese equities! For months, I've said that the valuations in Taiwan did not accurately reflect the fact that the "AI Hyperscale" runs (& will run) on TW hardware. It's the best performing segment of my AI winners list: citri.me/ai







