觅信聚合

我的关注

2026-09-01 00:37

原文：
Anthropic signs a $35B AI compute deal with $NVDA backed Lambda.

$HUT / DC, with Lambda / GPUs.

Frontier model demand is rapidly outpacing the ecosystem's ability to build capacity.

It also looks like Anthropic is securing more compute... Which neocloud/colo player is next? https://t.co/49yvFKjr49

翻译：
Anthropic 与 $NVDA 支持的 Lambda 签署了一项 350 亿美元的 AI 计算协议。

$HUT / DC，使用 Lambda / GPU。

前沿模型需求正迅速超出生态系统的产能建设能力。

看起来 Anthropic 也在获取更多计算资源……下一个会是哪个新云/机房玩家？ https://t.co/49yvFKjr49

2026-08-31 20:02

原文：
Interesting to see $QQQ spike when Trump gave a comment on potential rate hikes:

Answer: "Trump: No, I have a lot of respect for him, and he'll do what he has to do.

I think our interest rates are too high." https://t.co/vKwUWbUfHx

翻译：
有趣的是，当特朗普对潜在加息发表评论时，$QQQ出现了大幅上涨：

回答：“特朗普：不，我非常尊重他，他会做他必须做的事。

我认为我们的利率太高了。” https://t.co/vKwUWbUfHx

https://video.twimg.com/amplify_video/2094515156776886272/vid/avc1/1280x720/zu_gE1yuf7MOrOud.mp4?tag=29
Serenity (@aleabitoreddit) on X
Interesting to see $QQQ spike when Trump gave a comment on potential rate hikes:

Answer: "Trump: No, I have a lot of respect for him, and he'll do what he has to do.

I think our interest rates are too high."

2026-08-31 18:29

原文：
Macro has clearly cut high beta AI valuations over the past few months. But I try to stay focused on fundamentals.

If you're curious about my take though:

We're seeing record visibility with
- memory (2031), photonics (2029-2030), etc LTAs
- $NBIS, solana:5aQcoaSyy7fnhvpRHasBUAMH2nL5tTjVrmFw6BeGmoon, etc. massive demand imbalance for compute despite capex spend.
- combined with record projections ( $NVDA 70% growth 2027, $SMCI $60B+ new server orders, etc).

vs.

uncomfortable macro where:
- Warsh said the Fed has "more work to do" if underlying inflation isn't clearly returning to 2%.
- July PCE came in at 3.7% Y/Y
- July jobs report payrolls fell by 23k with unemployment at 4.1%
-10Y Treasury high/ 30Y highest since 2007
- Iran tensions continuing
- Japan Yen weakened to 160/ 2 rate hikes in Korea

And with rate 40/60 in favor of 25BPS near-term hikes.

There's definitely implications on financing + too many third/fourth order effects for me to personally map.

I'm not a macro expert but my take is:

The AI buildout will continue w/ $1.3T+ capex poured upstream expected next year, flowing through many bottlenecks and chokepoints.

While markets chopped off forward valuations and rotated toward currently profitable/defensive companies.

I don't know what's going to happen, especially since rate hike odds are almost a coinflip, and Trump posts can change sentiment overnight.

As long as I don't see indicators of hyperscalers cutting capex or demand wavering.

I'm staying long despite macro volatility, since I want ownership of the names controlling the future of the AI buildout.

And if there's any indicators of better macro climates, high beta are typically the fastest to recover.

翻译：
宏观因素显然在过去几个月削弱了高贝塔AI估值。但我仍尽量专注基本面。

如果你

2026-08-31 15:56

原文：
Mediatek + solana:LruuJRDNxyuVhpFe8eXK2keMFo9V9e4DzfERscBpump beneficiaries:

First thing that comes to mind is who benefits, if I had to rank from intial look:

1. Keystone Microtech (6683), $1.04B MC
Probe cards, SLT + test. Multiple Taiwan industry reports identify MediaTek as Keystone's #1 customer.

2. CHPT (6510), $2.8B
Advanced probe cards / (XPU wafer test)

3. Sigurd (6257), $3.3B
Wafer probing, burn-in, MediaTek backend test partner

4. WinWay (6515) $7.2B
test sockets/interfaces for MediaTek

5. KYEC (2449) $10.9B
testing, burn-in for MediaTek

Theres your Kinsus, Unimicron, Shennan Circuits, Zhen Ding, Tongfu Microelectronics, JCET, Ibiden, ASE, and other players that work with Mediatek as well (but less beta).

Airoha (6526) is probably the most interesting one i need to investigate more, since it's a Mediatek subsidary:

They do DSPs, TIA/optical connectivity, EML/SiPh/VSCEL drivers, SerDes, and others... where a lot of these happen to be large bottlenecks if you look at $SMTC + $AAOI earnings.

(don't own any of flagged names above, just sharing notes)

US related with Mediatek exposure would be:

$AMKR, $TER, $INTC, $TSM (yeah taiwan but us listed), $MU, then for design/IP prob $SNPS, $CDNS, $ARM.

But those are much less correlation (Mediatek is in Taiwan, so all your small suppliers are over there).

Again probably need to filter more into what benefits specifically with Mediatek's ASIC program

But usually a rising boat lifts all tides if they're suppliers in other segments.

Just initial look

翻译：
Mediatek + solana：LruuJRDNxyuVhpFe8eXK2keMFo9V9e4DzfERscBpump 受益者：

首先想到的是谁

2026-08-31 13:26

原文：
Yep, with $NVDA investing $3.5B into Mediatek.

It does feel like Nvidia is playing kingmaker with the next generation of ASIC winners like $MRVL / Mediatek, with a second order effect of eroding $AVGO position.

Playbook feels similar to what they did with Neoclouds like $NBIS and $CRWV (vs. hyperscaler cloud programs)

Help create the next leaders, then make them both financially and strategically aligned with Nvidia.

I do think it’s brilliant, but moves like this might push $AMD, $AVGO, and certain hyperscalers closer together.

Regardless this move cements Nvidia’s strategic position in inference (where hyperscaler ASICs were a former long-term bear case).

翻译：
是的，$NVDA 正在向 Mediatek 投资 35 亿美元。

感觉 Nvidia 正在为下一代 ASIC 胜者（如 $MRVL / Mediatek）充当王者制造者，同时产生二级效应，削弱 $AVGO 的地位。

其操作手法与他们在 Neoclouds（如 $NBIS 和 $CRWV）上的做法相似（相对于超大规模云计划）。

帮助培育下一代领袖，然后使它们在财务和战略上与 Nvidia 保持一致。

我认为这很聪明，但此类举动可能会让 $AMD、$AVGO 以及某些超大规模云厂商更趋于靠拢。

无论如何，此举巩固了 Nvidia 在推理领域的战略地位（此前超大规模云 ASIC 被视为长期空头）。

2026-08-31 12:43

原文：
Today, $NVDA has announced a $3.5B investment into Mediatek.

It appears Nvidia is "networking" itself into the broader ASIC ecosystem, from $MRVL to MediaTek.

And have future ecosystems aligned with Nvidia standards. (maybe to fend off $AVGO as well) https://t.co/okdGKS51U3

翻译：
今天，$NVDA 宣布向 Mediatek 投资 35 亿美元。

看起来 Nvidia 正在将自身“网络化”到更广泛的 ASIC 生态系统，从 $MRVL 到 MediaTek。

并让未来的生态系统遵循 Nvidia 标准。（也许是为了抵御 $AVGO） https://t.co/okdGKS51U3

2026-08-31 10:56

原文：
Remember my $SOI thesis anon? Soitec interview commentary today:

- 80% of capacity reservation agreements expected within 1-2 weeks, with more than 10 in total.

- $200M Photonics-SOI revenue now described as "absolutely a floor"

$TSM also sees SiPH going mainstream in 2027 with market share topping 50%+ of the optical transceiver market

If you hold a monopoly-like position over a chokepoint:
(UBS estimates Soitec has 95% share of the market for substrates underpinning silicon photonics chips)

Revenue usually flows through you.

翻译：
还记得我之前的 $SOI 论点吗，匿名？今天 Soitec 的采访评论：

- 预计在 1‑2 周内完成 80% 的产能预订协议，总数超过 10 份。

- $200M 光子‑SOI 收入现在被描述为“绝对的底部”。

$TSM 也认为 SiPH 将在 2027 年进入主流，市场份额将超过 50% 的光收发模块市场。

如果你在关键节点上拥有垄断式的地位：
（UBS 估计 Soitec 在支撑硅光子芯片的基板市场中占有 95% 的份额）

收入通常会流经你。

2026-08-31 10:04

原文：
“Memory is a bubble.”Explain this then, KOSPI bears:

Samsung has reportedly reserved ~70% of its memory production capacity through 2031 via LTAs with $NVDA, $MSFT, and $GOOGL as main counterparties.

Five years of demand visibility… https://t.co/GUGWTkXkMc

翻译：
“内存是泡沫。”那就解释一下吧，KOSPI 空头们：
三星已通过长期协议向 $NVDA、$MSFT 和 $GOOGL 等主要合作方预留了截至 2031 年约 70% 的内存产能。
五年的需求可视性… https://t.co/GUGWTkXkMc

2026-08-31 09:17

原文：
Imagine Celestial ($MRVL acquisition) was public.

The valuation would probably be astronomical in US markets (if I had to guess, $6-10B)

Because 2028 revenue projections were $500m (q4 annualized runrate), 2029 revenue projections were $1B.

And because Celestial is a core part of hyperscaler CPO programs.

Celestial's 2026 Q2 earnings would have been:

- Close to $0 revenue (Marvell said "revenue and earnings since the acquisition were not material")
- Losing -$12.5M/quarter ($50M/year)

Would people who knew what they were talking about looking at qualification cycle players say...

1. based on Celestial's Q2 P/S numbers and because it was losing -$12.5M a quarter... That it's a worthless meme stock?

Or

2. with CPO hyperscaler opportunities for 2028, it should be valued at $6-10B?

And I think that's a core cultural disagreement between US / EU cultures + markets.

If you knew how to evaluate qualification cycle players, which number actually matters? I'd argue overwhelmingly the latter.

翻译：
想象 Celestial（$MRVL 收购）是公开公司。

它的估值在美国市场可能会天文数字（如果要猜的话，60‑100亿美元）。

因为 2028 年的收入预测是 5亿美元（第四季度年化跑道），2029 年的收入预测是 10亿美元。

而且 Celestial 是超大规模云服务商 CPO 项目的核心部分。

Celestial 2026 年第二季度的收益本来会是：

- 收入几乎为 0（Marvell 表示“自收购以来的收入和收益并不重要”）
- 每季度亏损 1250万美元（全年 5000万美元）

了解情况的人在看资格赛周期的玩家时会说……

1. 根据 Celestial 第二季度的市销率以及它每季度亏损 1250万美元……它是毫无价值的 meme 股吗？

2. 考虑到 2028 年的 CPO 超大规模云服务商机会，它的估值应该在 60‑100亿美元？

我认为这反映了美欧文化+市场之间的核心分歧。

如果你懂得评估资格赛周期的玩家，哪个数字才真正重要？我倾向于后者，压倒性地更重要。

2026-08-31 06:08

原文：
$AAPL is now an important AI company apparently?

-> "OpenAI Hoarding Tens Of Thousands Of Apple Mac mini and Mac Studio" for reinforcement learning

-> Anthropic is renting Apple Mac mini units from $AMZN to complete similar AI tasks

I've never seen a company win so hard by doing nothing.

But whenever I see Apple + AI, I think of my beloved $RPI. Apparently Anthropic is working with Raspberry Pi now for their new model Hardware Standard, as of the 27th.

There's nothing yet about Anthropic/OpenAI buying Raspberry Pi devices for similar tasks, but maybe some OSINT researcher can do some digging.

翻译：
$AAPL 现在竟然成了重要的 AI 公司？

-> “OpenAI 囤积数万台 Apple Mac mini 和 Mac Studio” 用于强化学习

-> Anthropic 正在从 $AMZN 租用 Apple Mac mini 设备来完成类似的 AI 任务

我从未见过有公司什么都不做却能如此轻松获胜。

但每当我看到 Apple + AI，我就会想到我心爱的 $RPI。显然 Anthropic 已于 27 日开始与 Raspberry Pi 合作，推出他们的新模型 Hardware Standard。

目前还没有关于 Anthropic/OpenAI 为类似任务购买 Raspberry Pi 设备的消息，但或许有 OSINT 研究员可以深入挖掘。

2026-08-31 05:35

原文：
$TSM Vice President of Advanced Packaging:

"Foundries can scale up the optical engine, but the real bottlenecks to large scale deployment lie elsewhere...

Including lasers, optical fiber, fiber-optic connectors, and test" (UDN)

Just some bucket ideas:

Lasers: $LITE, $SIVE, $COHR, LandMark (3081), LuxNet, Sumitomo, Furukawa
Optical fiber / FAU: Foci, Browave, $GLW, Sumitomo
Fiber Optic / CPO Connectors: BizLink, $GLW , Fujikura, Nextronics, COXOC
CPO testing: MPI (6223), Advantest, $FORM, Msscorps, $KEYS, Chroma

(disclosure: exposure to theme above, individual names not named by TSMC)

There are others I like as well, such as ShunSin, which now has explicit TSMC COUPE exposure but probably belongs under a different category like packaging/integration.

eg. "Foxconn unit ShunSin confirms TSMC COUPE partnership"

I don't think there's anything too new vs. H1, aside from ongoing news (eg. FAU looks like there will have a demand imbalance in 2027, from a report last week).

But it's helpful to reiterate overall beneficiaries from time to time...

翻译：
$TSM 高级封装副总裁：

> “代工厂可以扩大光学引擎规模，但大规模部署的真正瓶颈在其他方面…
> 包括激光器、光纤、光纤连接器和测试”（UDN）

一些概念性名单：

激光器：$LITE、$SIVE、$COHR、LandMark (3081)、LuxNet、Sumitomo、Furukawa
光纤 / FAU：Foci、Browave、$GLW、Sumitomo
光纤 / CPO 连接器：BizLink、$GLW、Fujikura、Nextronics、COXOC
CPO 测试：MPI (6223)、Advantest、$FORM、Msscorps、$KEYS、Chroma

（披露：持有上述主题的敞口，具体公司未被 TSMC 点名）

我还看好其他公司，例如 ShunSin，虽已明确拥有 TSMC COUPE 敞口，但可能更适合归入封装/集成类。

例如：“Foxconn unit ShunSin confirms TSMC COUPE partnership”

相较 H1 并无太多新信息，除了一些持续的新闻（例如上周报告称 FAU 预计在 2027 年出现需求失衡）。

不时重申整体受益者仍有帮助。

2026-08-31 04:29

原文：
ESMT (3006) is 1.9x P/E ratio off July's earnings annualized.

Latest net income was: $109M net income/July, which is ~$1.3B annualized. With ~$395.9M net cash and with $249.6M in inventory.

The market cap less net cash and inventory is equivalent to:

~17 months of July earnings.

Where monetizing that inventory benefits from rising memory prices.

Their ASP hike expectations were +105–113% Q/Q last quarter from other analyst notes.

DDR2/DDR3, NOR Flash, SLC NAND, prices are expected to keep rising H2.

The main question is duration, but I personally think legacy DRAM/NAND shortages tightens throughout 2027. Others might disagree.

Other analysts like Huanan argue that PSMC wafer hikes are largely pass through (we'll find out H2).

Most people are finding companies with ~20 forward P/E ending 2028 to be attractive...

But 1.9x current runrate P/E seems like an anomaly to me personally.

That's my take on the company/legacy memory theme, I'll let others/markets decide if that's undervalued.

翻译：
ESMT（3006）的市盈率为1.9倍（按七月年化收益计算）。

最新净利润为：7月净利润1.09亿美元，年化约13亿美元。净现金约3.959亿美元，库存约2.496亿美元。

市值扣除净现金和库存后相当于：

约相当于 17 个月的七月收益。

变现这些库存将受益于记忆体价格上涨。

其他分析师的报告显示，他们对ASP的涨幅预期为上季度环比增长105%–113%。

DDR2/DDR3、NOR Flash、SLC NAND的价格预计在下半年继续上涨。

主要问题是持续时间，但我个人认为传统DRAM/NAND短缺将在2027年持续加剧。其他人可能持不同观点。

华南等分析师认为PSMC晶圆涨价大多会转嫁（下半年会见分晓）。

大多数人认为截至2028年，前瞻市盈率约20倍的公司更具吸引力……

但我个人觉得当前1.9倍的运行市盈率是个异常。

这就是我对该公司/传统存储器主题的看法，价值是否被低估就交给其他人或市场判断吧。

2026-08-30 18:29

原文：
I'm going to miss the days of Korean retail reacting heavily to $EWY / Samsung / $SKHY daily volatility.

New piece: "Korean day traders flee leveraged chip ETFs as regulatory curbs bite"

- Trading value in ETFs designed to deliver 2x daily returns of Samsung / SK Hynix has fallen to just 4% of its June peak

- The requirement follows earlier measures, including a higher minimum deposit, aimed at cooling speculative activity

- And then came the final boss... A 5 day trading course where:

"Investors must download a Windows-only program and spend at least one hour each day trading with virtual funds"

$MSFT Windows wiped out the Korean retail.

翻译：
我要怀念韩国散户对 $EWY / Samsung / $SKHY 日波动的强烈反应的日子。

新消息：“韩国日内交易者因监管限制而抛售杠杆芯片ETF”

- 旨在实现三星 / SK Hynix 日收益 2 倍的ETF交易额已跌至6月峰值的仅4%

- 此要求是继此前包括提高最低保证金等措施之后，旨在抑制投机活动

- 随后出现了终极限制……一个为期5天的交易课程，要求：

“投资者必须下载仅限 Windows 的程序，并每天使用虚拟资金交易至少一小时”

$MSFT Windows 把韩国散户扫荡一空。

2026-08-30 15:24

原文：
The more I look at ESMT, the more excited I get researching legacy memory themes.

So looked through H1 sellside models...

From March ESMT reports:
Huanan: DDR3 4Gb ASP +50% QoQ in Q2
Fubon: cited roughly +40% Q2 DRAM pricing

By August... Huanan was citing ESMT Q2 ASP growth of:
+105–113% Q/Q blended.

Saying DDR3 pricing had come in substantially better than expected.

So legacy memory pricing translated into much more operating leverage at ESMT.

About PSMC wafer cost hikes: Huanan estimated manufacturing costs could rise ~100% in 2H26, but customers remained highly willing to accept the pass-through

If we look at July net income was roughly $109M... way above the earnings levels the earlier models were built around.

If I had to give a TLDR "consensus" of reports

1. DDR3/DDR2 supply remains structurally tight as competitors move capacity toward higher-value memory.
2. Concord saw no conditions supporting a price reversal during 2026.
3. Concord expected contract-price increases to cover higher foundry costs.
4. Huanan later found that customer willingness to accept those cost increases was stronger than expected.

Fun thing I read from Huanan was DDR3 is used in products such as IP cameras and HDDs where the memory component is a relatively small part of the customer's total BOM.

So even very large legacy DRAM price hike may only add a few dollars to the finished product's BOM... and for many customers, paying a few extra dollars would be more economical than redesigning + requalifying (but this flows materially into ESMT's net income)

Anyway, just excited about this idea (*disclosure I have positions)

TLDR:
-> Huanan updated ESMT blended Q2 ASP to +105-113% Q/Q
-> Given July's blowout net income (putting ESMT at 1.9x p/e annualized RR), implications for the future are exciting if inelasticity continues + ESMT has the ability to hike further due

2026-08-30 15:24

to low % of finished product BOM cost.
-> Indication customers were still willing to accept pass through for PSMC wafer hikes.

翻译：
我越是研究 ESMT，就越对传统内存主题感到兴奋。

于是查看了 H1 的卖方模型…

来自 3 月 ESMT 报告：
Huanan：DDR3 4Gb ASP 环比 Q2 增长 +50%
Fubon：提到 Q2 DRAM 价格大约上涨 +40%

到 8 月… Huanan 报告 ESMT Q2 ASP 综合增长：
+105–113% 环比。

称 DDR3 价格远好于预期。

因此传统内存定价为 ESMT 带来了更高的经营杠杆。

关于 PSMC 晶圆成本上涨：Huanan 估计制造成本在 2H26 可能上升约 100%，但客户仍高度愿意接受成本转嫁。

如果我们看 7 月净利润大约为 $109M… 远高于早期模型所依据的盈利水平。

如果要给出报告的 TLDR “共识

2026-08-30 11:41

原文：
Just some random ideas from Elon’s post:

$AAON - Chillers / liquid cooling
$NVT - Cooling
$MOD - Chillers / Cooling
$IESC - Wiring
$MTRS - chillers
$JCI - chillers
$HPS.A - Transformers (disclosure I own positions in this)
$PRY.MI - networking/wiring(fiber)
$CRDO / $AAOI (same) / $LITE / $COHR - networking

Since Elon Musk cited:

- transformers
- wiring
- liquid cooling
- massive chillers + complex networking

As the point of failure for AI compute buildout.

Which is “harder than finding power”

翻译：
以下是 Elon 帖子中的一些随意想法：

$AAON - 冷水机 / 液体冷却
$NVT - 冷却
$MOD - 冷水机 / 冷却
$IESC - 布线
$MTRS - 冷水机
$JCI - 冷水机
$HPS.A - 变压器（披露：我持有该仓位）
$PRY.MI - 网络/布线（光纤）
$CRDO / $AAOI（相同）/ $LITE / $COHR - 网络

自 Elon Musk 提到：

- 变压器
- 布线
- 液体冷却
- 大型冷水机 + 复杂网络

作为 AI 计算扩展的故障点。

这比寻找电力更难。

2026-08-30 11:08

原文：
FYI to the AI bears: $NVDA projects $1.3T of hyperscaler spend for 2027.

Up from ~$800B in 2026, alongside ~70% revenue growth despite remaining capacity constrained… far above prior Street expectations of ~45%.

That capex flows across the AI supply chain… from memory and networking to foundries. Especially through the new bottlenecks/chokepoints that didn’t benefit from older generations.

We’re like 2 hours into the party where people started off drinking Sapporo draft beer.

Someone spiked Situational’s drinks and we had to pause the party for 30 min.

But now it’s back on with tequila shots on the table.

I think the real party is about to begin.

翻译：
给AI空头们的提醒：$NVDA 预计2027年的超大规模算力支出将达 $1.3T。

相比2026年的约 $800B 大幅上升，同时收入增长约 ~70%，尽管产能仍受限……远超此前华尔街约 ~45% 的预期。

这笔资本支出遍布AI供应链……从内存、网络到代工厂。尤其是那些新出现的瓶颈/卡点，旧一代产品并未受益。

我们现在已经在派对上喝了大约 2 小时，大家最初在喝 Sapporo 生啤。

有人在 Situational 的饮料里下了药，我们被迫暂停了 30 分钟。

但现在派对重新开始，桌上摆着 tequila shots。

我觉得真正的狂欢即将开启。

2026-08-30 09:05

relative to future qualifications/partner size/capacity/potential.

$SIVE needs to position themselves as a forward looking, global hypergrowth optical company supplying lasers to hyperscaler programs.

And not let the narrative get dominated by backward looking metrics.

And as Morgan Stanley put it... "Key [CPO] participants include … $LITE, $COHR, and Sivers laser supply".

I'm personally a happy $SIVE shareholder for high-beta exposure to the next 2027-2028 optical shift with 1.6T/CPO.

翻译：
我对 $SIVE 从财报电话会得出的最终总结：

2027 年可插拔方案（1.6T 放量）：

- $JBL（1.6T LRO，已确认）2027 年上半年下单，2027 年下半年放量
-> -> 大概率有多个超大规模云厂商客户

- 3 家可插拔方案厂商（评估中）
- 3 家可插拔方案厂商（供应评估中）

面向 2027 年下半年至 2028 年及以后的参考设计：

- $GFS（Globalfoundries）SCALE（已确认）
-> -> 大概率 $AMD 以及其他超大规模云厂商会成为 CPO 客户

2027 年的外置光源 / 光引擎：

- $POET（已确认）
-> -> Lumilens / 超大规模云厂商终端用户

- O-Net（已确认）
-> -> 面向超大规模云厂商和亚洲市场的大型 ODM/OEM

- 大概率 $AEVA（在映射中）
-> -> 向 OE 供货，再到超大规模云厂商 / 激光雷达

- SemiNex（已确认）

2028 年的光 I/O：

- Ayar Labs（已确认）
-> -> 大概率 $AMD、ALchip，以及超大规模云厂商 ASIC 项目终端用户 + $NVDA NVLink

- $MRVL Celestial（2023-2024 年披露中的潜在客户）
- Lightmatter（2023-2024 年披露中的潜在客户）
- Lightelligence（2023-2024 年披露中的潜在客户）

另外，2028 年也可能有 $AAPL，用于下一代可穿戴更新，以及与 Lightium 的 TFLN 相关的其他项目。

在产能分配方面：
- Win Semi
- 另一家晶圆厂（分配量非常大）

当时正处于 InP 连续波 DFB 激光器短缺。

这些都是 Sivers 正在推进的持续开发/认证，因为其目标是 2027-2028 年的下一代 SiPH + CW 1.6T 和 CPO。

尤其是高盛估算 CPO 机会会从如今几乎接近 0，增长到 2028 年约 910 亿美元 TAM。

不过我的批评是：

=> 核心重点并没有放在向西方受众传达 2027-2028 年光学放量的经济规模。
=> 关于纳斯达克上市的法律式模糊表述应该删掉，并且需要更明确的方向，更快执行。
=> 过多精力放在为当前较小的收入/TTM 收入/管线转化（allspace、Tachyon）辩护，而不是强调未来的认证、伙伴规模、产能和潜力。

$SIVE 需要把自己定位成一家面向未来、全球化的高增长光学公司，为超大规模云厂商项目供应激光器。

而不是让叙事被落后指标主导。

正如摩根士丹利所说……“关键 [CPO] 参与者包括 … $LITE、$COHR，以及 Sivers 激光供应”。

我个人很看好持有 $SIVE，作为押注 2027-2028 年 1.6T/CPO 光学变革的高贝塔敞口。

2026-08-30 09:05

原文：
Just my final recap on thoughts + fundamentals of $SIVE from ER:

Pluggables (1.6T Ramp) for 2027:

- $JBL (1.6T LRO, confirmed) Orders H1 2027, Ramp H2 2027
-> -> Likely multiple hyperscaler customers

- 3 pluggable makers (evaluation)
- 3 pluggable makers (supply assessment)

Reference Design for likely H2 2027–28+:

- $GFS (Globalfoundries) SCALE (confirmed))
-> -> Likely $AMD and other hyperscalers as CPO customers

External Light Sources / Optical Engines for 2027:

- $POET (confirmed)
-> -> Lumilens / hyperscaler end users

- O-Net (confirmed)
-> -> Massive ODM/OEM for Hyperscalers + Asia

- Likely $AEVA (mapping)
-> -> supplying to OE -> hyperscalers / lidar

- SemiNex (confirmed)

Optical I/O for 2028:

- Ayar Labs (confirmed)
-> -> Likely $AMD, ALchip, and hyperscaler ASIC program end users + $NVDA NVLink

- $MRVL Celestial (potential customer from 2023-2024 disclosures)
- Lightmatter (potential customer from 2023-2024 disclosures)
- Lightelligence (potential customer from 2023-2024 disclosures)

Then potentially $AAPL for 2028 for next-generation wearable updates and other programs from TFLN with Lightium.

In terms of capacity allocations:
- Win Semi
- 1 other foundry (with tremendous allocations)

during an InP CW DFB laser shortage.

These are all on-going developments/qualifications since Sivers is targeting next-generation SiPH + CW with 1.6T and CPO for 2027-2028.

Especially as Goldman Sachs models the CPO opportunity going from effectively near-zero today to ~$91B TAM by 2028.

However my criticisms were:

=> Main focus was not on communicating economic scale of 2027-2028 optical ramps to Western audiences.
=> Legal vagueposts around NASDAQ listing should be dropped, and clear direction should be set + executed faster on.
=>Too much focus was put on defending smaller current revenue/TTM revenue/pipeline conversion (allspace, Tachyon.)

2026-08-30 06:24

thesis and reasoning so everyone can learn from it.

Than one where information increasingly disappears behind $100 paywalls or becomes diluted with engagement-bait noise.

Hopefully I can help create a new culture on X rather than amplify the old one.

翻译：
看到我在 X 上接触过的一些人身上有个模式，多少有点失望：

- 试图和同一批受众互动，然后很快转向激进变现

- 当我选择不参与或不推广付费社群时就很不高兴

- 抛出人身攻击或不讲逻辑的攻击，还想把这说成只是对基本面的不同意见

我刚加入 X 的时候，很多金融内容都围绕着手表/私人飞机 + 期权、号称能提供“华尔街秘密”的付费社群，以及一些乱线图表。

我一直想做得不一样的一点，就是把我的研究和逻辑公开发布，给所有人看。

我觉得这确实引起了很多人的共鸣，毕竟我的粉丝增长就是证明。

但当然，这种增长也会和一些人的商业模式产生摩擦，他们靠把 X 当成引流工具，把受众导向付费服务或昂贵的付费墙。

我的订阅费几个月来一直是 $1，而且还会保持在最低档，尽管一开始有人说我会涨到 $100。

我所有核心论点都已经免费发布在主时间线里了。我完全可以把更多内容放到付费墙后面。但我就是不想。

我只是更喜欢公开分享想法，让每个人自己判断。

很多想法都进展得很好：$AXTI、$NBIS、$AEHR、$MU、$INTC、$EWY、$MRVL、$LITE、$RPI、$IQE、$SOI、$TSEM、$ARM、$COHR，还有很多更多。

也有一些因为行业意外延迟而受阻，比如用玻璃基板的 $LPKF 和 Auros 的混合键合。

还有一些会一直延续到 2027/2028 年，包括 $SIVE、Foci、Shunsin、$CCXI、$XFAB 以及更广泛的 CPO/人形机器人主题。

也有一些是全新的，比如 Etron/ESMT 的 DDR2/DDR3。

还有很多像 $JBL 到 Murata 再到 $MXL，我只是随便分享，看看这个想法对不对。

不是每个想法都会对。这就是研究的一部分。

但我更希望 X 上是一个大家公开发布论点和逻辑、让所有人都能从中学习的平台。

而不是一个信息越来越消失在 $100 付费墙后面，或者被为了互动而造的噪音不断稀释的平台。

希望我能帮助在 X 上创造一种新文化，而不是放大旧文化。

Loading...