Evan老曹

7年战略投资、3年TPM经验 / 用数据、财报和商业逻辑，拆解前沿科技商业议题，看懂趋势和机会 / 微信公众号：Evan老曹

[Evan老曹 @evancaojt]

Gemini 3.8 Falsh 发布了！没人说吗？？？

[Evan老曹 @evancaojt]

造火箭的SpaceX跨界造燃机叶片？马斯克到底在下什么大棋？谁可能受益？ https://t.co/rcaXLnMJ62 来自 @YouTube

[Evan老曹 @evancaojt]

Notion AI 里面的 Fable 5 是真难用，又慢又傻。

一个简单任务在运行1个多小时后���我不得不中止，不然我担心这个月的额度被这个任务消耗完。

真是失望！

[Shay Boloor @StockSavvyShay]

Cathie Wood bought ~$53M of $NVDA today

[Evan老曹 @evancaojt]

https://t.co/pOFWCz8x1W

[Evan老曹 @evancaojt]

宇树中一签挣47.46万元。
有中签的吗？

[Evan老曹 @evancaojt]

NeoCloud（新云服务商）最近被推到了聚光灯下。

简单说，这些公司专门提供GPU算力，服务AI训练和推理，而不是传统的通用云计算。它们中的不少，前身是加密货币矿工。

为什么现在应该多看一眼？

先看最突出的例子——CoreWeave。它2017年起步时还在挖以太坊，2019年转型做GPU云。结果AI浪潮来了，它手里的电力合同、数据中心和GPU库存，瞬间变成稀缺资源。a16z的数据显示：CoreWeave才跑到第25个季度，季度营收就已经超过当年Azure、AWS、Google Cloud在第30个季度时的水平。增长曲线近乎垂直拉升，速度远超早期云巨头。

这背后的逻辑其实很直白：AI最缺的不是算法，而是��、机房和高端GPU。传统超大规模云扩容跟不上，而前矿工们早就把最难拿的电力接入权和高密度机房布局搞定了。再加上和NVIDIA的紧密关系，它们成了AI算力市场的“中间层”——既比巨头更灵活，又比纯创业公司更有落地能力。

目前公开市场里，CoreWeave、Nebius、Applied Digital等几家已经形成了相对清晰的梯队。CoreWeave凭借规模和客户质量（微软、OpenAI、Meta、Anthropic等）明显领先，合同积压动辄上千亿美元。其他公司则更多在区域电力或特定客户上发力。

这波行情的吸引力在于：它是AI基建里少有的、已经能看到真实营收高速增长的细分赛道。不是概念，而是实打实的订单和电力落地。

但必须同时看清风险。这些公司是典型的重资产、高杠杆生意。买卡、建机房、拉电力全是烧钱，资本支出往往比营收涨得还快。GPU折旧成本极高，债务利息也不低。目前多数公司仍在亏损或利润率极薄，真正的自由现金流还要看后续几年能否把利用率做上去、把融资成本降下来。

这是一场“用高杠杆赌AI算力长期需求”的游戏。需求如果持续爆发，它们会继续高增长；一旦供需阶段性缓和，或者新一代芯片大幅降价，高固定成本和债务就会成为负担。所以，值得关注，不等于现在就盲目重仓追高。更务实的做法是：把NeoCloud当成AI基建板块的一个观察窗口，而不是单一押注。

��结一句：NeoCloud的崛起，本质是“旧资产遇上新需求”。它们用挖矿时代积累的电力和GPU优势，踩中了AI最紧的瓶颈。增长真实，机会存在，但重资产属性决定了波动会很大。普通投资者应该先把逻辑看清楚，再决定自己的参与方式。

关于NeoCloud新云服务商，我计划专门制作一期视频拆解它们的机会和风险。感兴趣的朋友们，可以关注我的B站/Youtube（搜索：Evan老曹）。

[Evan老曹 @evancaojt]

热知识：中国居民储蓄率约 32–35%。

[Global Markets Investor @GlobalMktObserv]

🤯THIS IS MIND-BLOWING:

US personal savings rate is down to 2.7%, the lowest since June 2022.

Excluding 3 months in 2022, this is the lowest level of savings since the Great Financial Crisis.

Since the start of 2024, savings have more than HALVED.

To put this into perspective, before this century, the savings rate never fell below 3.5%.

This comes as elevated inflation translates into higher living costs, while elevated borrowing costs erode consumers' ability to save.

The US consumer may be closer to the edge than it looks.

[Evan老曹 @evancaojt]

看看它们的的财报吧，新云服务商确实值得关注！

我的下一期视频选题就它了。

[Melvin @MelvinInvests]

JPMorgan is now forecasting that the seven biggest hyperscalers will spend a combined $1.489 trillion on Capex in 2027 .

That figure is up from $901 billion in 2026 and just $443 billion in 2025, which means spending has more than tripled in just two years .

The growth rate is still running at 65% year over year even after cooling from 2026's explosive 103% pace .

JPMorgan separately projects that total global AI and data center investment could reach $5 trillion to $7 trillion by 2030, which shows this hyperscaler figure is really just the visible tip of a much larger buildout.

Now here is how you can benefit from all of this because none of that trillion plus dollars goes only toward chips, since a massive share flows to the neoclouds, power suppliers and physical infrastructure builders that make the entire buildout possible.

The neocloud layer is where some of the fastest growth in the entire market is happening right now.

CoreWeave, already named directly in JPMorgan's forecast, saw its backlog swell to $130 billion, up from $99 billion just one quarter earlier, and the company raised its full year guidance after a blowout print.

Nebius reported second quarter revenue of $582 million, up 454% year over year, and its stock jumped more than 25% after the print, with the company now holding over $40 billion in customer commitments backed by anchor contracts from Meta and Microsoft.

IREN signed $2.8 billion in new multiyear cloud contracts and raised its year end annualized run rate target to $4 billion.

Grid equipment makers are seeing the sharpest pricing power further down the chain, since transformer lead times have stretched to two to four years and prices have risen roughly 80% over the past five years.

Vertiv has returned roughly 69% year to date on a $15 billion order backlog built from 252% order growth and the company supplies the liquid cooling and power systems that go inside every new facility.

Hubbell and nVent Electric round out the grid hardware group, since both companies supply the transformers, enclosures, and connectors that hyperscalers need to build these facilities.

Electrical construction firms are converting these capex announcements directly into multi year backlogs.

Comfort Systems has returned nearly 88% year to date doing HVAC and electrical work specifically for data centers, while Sterling Construction is up over 122% and Quanta Services has climbed 56%.

EMCOR Group and MasTec are doing similar specialty electrical and mechanical work, riding comparable order backlogs.

Power generation names are capturing a piece of this spending too, since data centers can't wait years for new grid connections to come online.

Caterpillar's Power Generation segment grew 41% for four straight quarters supplying on site backup turbines, while GE Vernova booked $2.4 billion in data center equipment orders in a single quarter, more than its entire prior-year total.

Bloom Energy, Constellation Energy, Vistra, and Talen Energy are all locking in long-term power contracts directly with hyperscalers and neoclouds.

$1.5 trillion in hyperscaler capex has to flow somewhere and I'm tracking every layer it touches from neoclouds and power to cooling, grid equipment, and construction.

The AI infrastructure buildout is just getting started, make sure to follow @MelvinInvests for more AI infrastructure insights and if you want to see exactly what I'm buying as an analyst at Milk Road Pro, you can come join us for just a $1. https://t.co/Hkywss4Ugh

[Evan老曹 @evancaojt]

这简直离谱，��又很常见。

[First Squawk @FirstSquawk]

ANTHROPIC PROJECTS $190 BLN-$200 BLN IN REVENUE FOR 2028, ACCORDING TO SOURCES, WITH WALL STREET USING 2028 REVENUE AND AN EV-TO-REVENUE MULTIPLE TO VALUE THE IPO

[Evan老曹 @evancaojt]

InP这波短缺比预期狠多了。

衬底从去年Q4起已经连涨三次，现在直接进���四连涨；外延片也要冲第三涨。Q4单次涨幅要破10%，供应商原话是“有钱也未必买得到”。

核心卡点还是上游：政策+产能双重受限，全球有效供给大概只能覆盖30%需求。AI数据中心从铜光混合往更高光学比例切，800G/1.6T对InP激光芯片的消耗是倍数级的，这东西短期根本扩不出来。
光学互连正在成为继GPU、HBM、电力之后，AI基础设施下一个真正硬瓶颈。

后面光模块成本传导和部署节奏，值得持续盯。

[P Equity Research 📰 @pequityresearch]

InP shortage is getting worse as expected.

> A further price increase is anticipated in the fourth quarter, with the increase expected to exceed 10%, marking the largest increase in history. Suppliers have stated that "even with money, you may not be able to buy them."

> Industry insiders revealed that the price of indium phosphide substrates started to rise in the fourth quarter of last year and has been raised three times since then, now reaching a "fourth consecutive increase"; the epitaxial wafers made from the substrates have also increased in price twice, and are heading towards a "third consecutive increase" in the fourth quarter, which means that the market supply shortage is more severe than expected.

> Indium phosphide (IP) epitaxial wafers must be manufactured from IPP substrates, but upstream IPP substrate suppliers are constrained by policies and production capacity, and have so far been unable to meet market demand. Due to this supply-demand imbalance, prices have been rising steadily, with increases from the initial 3% to 5%, and the latest news indicates a potential surge of over 10%, the largest increase in history.

We need more InP...

[Evan老曹 @evancaojt]

@pequityresearch @equity_neconeco AXT的生产主体却在中国

[Evan老曹 @evancaojt]

@aleabitoreddit 很有意思的分享。

好的投资者：
>忘记了自己的账户
>没有积极交易/过度交易（不频繁交易者18.5%的回报率 vs. 11.4%）

[Serenity @aleabitoreddit]

Now that markets are hosting a laser party again from $SIVE, OE Solutions, $LITE, Coherent, to $AAOI.

There’s a pretty interesting study:

Both from Fidelity and a UC Berkeley research paper, that the best investors are the ones who either…

- Anecdotally forgot about their account (Fidelity)
- Didn’t actively trade/overtrade (18.5% return from infrequent traders vs. 11.4%).

Not any advice, but some of these anecdotes might be helpful to retail to read in general…

Since I witnessed a lot capitulation off memory, photonics, or thematic volatility, just for retail not have positions on a sharp recovery.

Having conviction also usually people in that “not overtrading” camp, since it helps to not overtrade in drops or see opportunities to cost average.

[Evan老曹 @evancaojt]

抄袭都这么明目张胆了吗？真是恶心。

[Evan老曹 @evancaojt]

我这样的还能抢救一下吗

[Evan老曹 @evancaojt]

@aleabitoreddit https://t.co/XwdVqVGw8p

[Evan老曹 @evancaojt]

停留一分钟。

看懂CPO的构成，是投资它的第一步👇

[Evan老曹 @evancaojt]

停留一分钟。

看懂CPO的构成，是投资它的第一步👇

[Evan老曹 @evancaojt]

一分钟看懂CPO构成。

GPU越来越多，交换机/服务器节点间的数据传输成了瓶颈。

传统光模块虽能满足当前需求，但同样的走线距离，在800G、1.6T速率下损耗和功耗急剧上升。

产业转向新方案——CPO（Co-Packaged Optics，共封装光学）。

简单来说，CPO就是把光引擎与交换芯片共同封装在同一基板/中介层上，同等速率下更省电、带宽密度更高。

>CPO由哪些部分构成？

CPO��新设计了芯片与光的连接方式：

1. 交换芯片：核心大脑。 负责服务器间数据交换，如英伟达、博通的AI网络交换芯片。

2. 光引擎：CPO的核心。 由调制器（把数据“写”到光上）、光探测器（光→电）、驱动/TIA和光波导组成，不含激光器。

3. 硅光芯片：光引擎的核心技术路线。 用半导体工艺，把光波导、调制器、探测器集成到硅芯片上，更集成、更低成本、更适合大规模制造。它与光引擎的区别：光引擎是完整模块，硅光芯片是核心芯片；光引擎不一定用硅光，硅光最主流。

4. 激光源：外置为主。 主流是外置激光源（ELS）：激光器提供连续光（CW），经保偏光纤从面板送入光引擎；更易维护、寿命问题更好解决。

>CPO对光模块有什么影响？

位置变了。 传统：交换芯片→PCB电路→光模块→光纤；CPO：光引擎与交换芯片共同封装，光纤直接接入。只“放附近”叫NPO，共同封装才是CPO。

需求可能下降。 传统光模块含激光器、PCB、DSP、光接口、外壳；CPO中部分功能被拆分进光引擎，可插拔光模块的价值量可能下降。

产业链变化。 传统：激光器→光器件→光模块厂商→交换机；CPO：激光器→硅光芯片→光引擎→交换芯片封装。价值转向硅光芯片、光引擎、封装。

不会马上取代。 光模块成熟、可插拔、维护方便；CPO散热复杂、维修难、生态待建。未来可能长期共存：普通数据中心用传统光模块，超大规模AI集群用CPO。

>总结

CPO本质是把光引擎与交换芯片共同封装在同一基板上，缩短板上电信号传输距离，提升AI集群互连的能效和带宽密度。它解决交换机/服务器节点间互连，不是芯片内部的光I/O。核心组成：交换芯片、光引��、硅光芯片、外置激光源。对光模块行业，CPO不是替代，而是升级；AI时代竞争不只是算力，更是芯片+存储+光互联能力。

如果你喜欢我的内容，关注获取更多拆解。

[Evan老曹 @evancaojt]

@crytogamefi47 越南的朋友们，很开心认识你们

[Evan老曹 @evancaojt]

@mayxhen 借楼郑重说一声，我有关必回，讲个就是个诚信

[Evan老曹 @evancaojt]

有没有过来人讲讲，你们是怎么把账号做起来的？

[Evan老曹 @evancaojt]

小伙伴们，大家好。

我发现在X上做内容，不能光埋头苦干，好内容没人看是瞎折腾。

要多交朋友，互相多支持。

留下你的声音，我会成为你的朋友。

希望明早醒来，我的朋友圈能突破400。