Bit & Byte

[Bit & Byte @giga_bit_]

200 x 300 days/day =60,000 per yr

‘26 100K robotaxi $6B -> 1T
‘27 300K robotaxi $18B -> 1.6T
‘28 800K robotaxi $48B -> 2.4T
‘29 1200K robotaxi $72B -> 4.2T
‘30 2000K robotaxi $120B -> 6.9T

[Serenity @aleabitoreddit]

Does anyone else have this annoying seed planted in their mind too?

Whenever I see $MRVL drop, I get reminded of Jensen’s comment “The next $1T company”.

[TheValueist @TheValueist]

$GLW Corning: The AI Optical Infrastructure Springboard Strategy. Investment Thesis. New: 6/29/26.

Corning has repositioned a traditional materials manufacturer into a critical supplier at the physical layer of AI infrastructure — a transition that leverages decades of fiber optics and photonics expertise into a market where that expertise is now mission-critical rather than commoditized. High-density fiber and advanced photonics are the connectivity backbone that AI data centers require as networking bottlenecks become increasingly binding constraints on cluster performance, and Corning's manufacturing scale and materials science depth position it as one of the few suppliers capable of meeting that demand at the volumes hyperscalers require.

The NVIDIA and Meta partnerships are significant validation points. Both represent demanding customers whose technical requirements and qualification standards are rigorous, and their direct engagement with Corning signals that the company's optical infrastructure capabilities are viewed as strategically necessary rather than substitutable. That kind of direct hyperscaler relationship is difficult for competitors to displace once established, given the integration depth required in data center network architecture planning.

The Springboard strategy is the financial framework converting the AI infrastructure opportunity into demonstrated margin improvement. Operating margin expansion that has already materialized provides credibility to the more aggressive 2030 revenue growth targets — this is not a purely forward-looking narrative but one with a track record of execution behind it. That said, the distance between current results and the 2030 targets is substantial, and the growth trajectory assumes continued AI infrastructure capital spending at a pace that has historically been difficult to sustain without periodic digestion phases.

Capital intensity is the structural constraint on returns during the buildout phase. Scaling fiber and photonics manufacturing capacity to meet AI-driven demand requires sustained capital deployment, and the return on that investment depends on demand durability matching the capacity being built. Customer concentration compounds that risk — significant revenue exposure to a small number of hyperscaler relationships means that any shift in AI infrastructure capital spending plans at a major customer would disproportionately affect Corning's growth trajectory relative to a more diversified customer base.

The Solar business is a complicating factor that sits somewhat apart from the core AI optical infrastructure narrative. Scaling that segment successfully requires different operational capabilities and serves a different demand driver, and management attention split across a capital-intensive solar scale-up alongside the AI infrastructure buildout introduces execution complexity that pure-play AI infrastructure companies don't carry.

The valuation reflects multiple years of anticipated growth, which means the premium is justified only if execution stays on pace with the Springboard targets and AI infrastructure capital spending remains robust through the multi-year buildout period the thesis depends on. Corning's market position is genuinely dominant in its core optical infrastructure niche — the question is whether that dominance, expressed through a still-developing financial trajectory, supports a price that has already captured much of the anticipated upside.

[川沐｜Trumoo🐮 @xiaomustock]

想到一个问题，这几天看到英伟达190低估，20倍市盈率，按照每年几乎100%的增速，明年就是10倍市盈率。但是英伟达想实现100%增速，那么就要搭配double的hbm，那么就需要双倍的消费级dram去造hbm，但是明年存储还是没有新增产能，这个局面最终会怎么样？
我猜测可能路径是明年某个时刻三星海力士找长鑫进口dram，然后再在自己家工厂封装成hbm4e。因为英伟达不可能直接用长鑫造的hbm，技术和国情都不允许。

[Bit & Byte @giga_bit_]

感觉年底冲500 都不是问题！

[美研芒格君 @Kay2289123]

$MRVL oh，这篇 MRVL (Marvell) 的文章获得了比我想象中更多的���注，已经 270K 浏览了，我真的建议每一个要在 Marvell 上投资的一美元人都至少看一下这篇文章

MRVL（Marvell），我已经关注了将近三年，曾在起始的时候大力加仓，在大多数人最不看好的时候坚定持有。这篇文章的写作花了我 40 小时，

这篇文章浓缩了我大部分对它现浓缩了我大部分今年五月对 Marvell 的认知精华。如果说唯一的认知改变，就是它从我之前认为的定制芯片公司，已经全面变成了互联公司

其中包括但不限于：
1. 算力互联
2. 内存互联
(a) Scale up
(b) Scale cross
(c) Scale out

我会持续关注 MRVL，它是我的第一大追踪

[美研芒格君 @Kay2289123]

$MRVL oh，这篇 MRVL (Marvell) 的文章获得了比我想象中更多的关注，已经 270K 浏览了，我真的建议每一个要在 Marvell 上投资的一美元人都至少看一下这篇文章

MRVL（Marvell），我已经关注了将近三年，曾在起始的时候大力加仓，在大多数人最不看好的时候坚定持有。这篇文章的写作花了我 40 小时，

这篇文章浓缩了我大部分对它现浓缩了我大部分今年五月对 Marvell 的认知精华。如果说唯一的认知改变，就是它从我之前认为的定制芯片公司���已经全面变成了互联公司

其中包括但不限于：
1. 算力互联
2. 内存互联
(a) Scale up
(b) Scale cross
(c) Scale out

我会持续关注 MRVL，它是我的第一大追踪

[美研芒格君 @Kay2289123]

https://t.co/3TWZzaX5p4

[川沐｜Trumoo🐮 @xiaomustock]

今天也算是半年过去，对自己上半年的惨痛交易也做个总结，
1.尽量买标品生意能流水线生产的或者互联网能直接复制产品的公司。
诺基亚这一俩个月俩回没啥起色，可能就跟它非标品的生意模式有关，基站啥的只能靠人力一点点干，搞土地搞关系搞人力搞供应链，苦力生意。
2.尽量买护城河极高，能参与公司极少的垄断公司垄断行业，湖南裕能这些就是对手多，无脑扩产的太疯狂，很难形成自己稳固的定价权。
3.自己没法掌控它实时仓过往业绩也不是特��好的etf或者lof少碰。当时5月初选港美互联网lof这个标是因为它持有美光和闪迪。如果基金经理完全按照一季度持仓加仓，则后续理论上表现会非常夸张，毕竟美光和闪迪在6月都涨了快50%，但真想不通这个傻逼经理在秀什么几把操作，后续价格走势完全跟美光和闪迪无关联。简直是盲盒交易。远不及中韩半导体这种etf持仓稳固，几个季度过去一直高比例持有三星海力士。
4.最重要最严重的错误，融资上杠杆。
差点在3月份爆仓死掉，2x海力士和ewy都暴跌差点没熬过去。以后再不去买超过自己本金的资产，顶多在自己本金范围内购买2x扩大上升周期的收益。
最沉痛的教训，当时3月的暴跌不仅吞��了我1/2月份的收益，还直接巨亏打到了接近失去上桌的机会。
后续市场肯定还会有类似的波动，
比如真的加息来临，
比如市场承受不住大把的AI公司发债融资，
比如存储高价可能大幅抑制消费端需求，造成需求减弱，或者存储后期大幅度扩产引发降价等等。
总之市场会在我们意想不到的方式来一个黑天鹅来摧毁杠杆投资者，一俩个月风声过去，股市继续歌舞升平，全场新高，但与你无关。

[CapitalFlow @CapitalFlowApp]

Whales dump $17.2M in $NRG calls⁠ as spiking summer power grid demands trigger massive profit-taking.

[川沐｜Trumoo🐮 @xiaomustock]

按照 $GOOG 财报会中透露的时间节点，26年底tpu出货，27年放量。
对应的 $MRVL 连续爆拉时间可能就在9月左右开始直到27年。因为tpu没有大规模放量现在市场看不太清楚mrvl的潜力。

[0xkevin (🖤 , 💙) @0xKevin00]

“捏住AI的蛋蛋”，快来抄川沐大神的股票作业

川沐公开喊单的票很少，去年 $150 的 $MU ，4月 $50 的 $MRVL 等等

现在他公开喊单三只票：$sk $glw $ASX ，一套完整的AI基础设施利润链

川沐的核心投资框架可以用一句话概括：AI产业链里谁赚钱最确定、最多，就买谁，不碰不确定的

所以这三只标的不是随机选的，而是从整条AI产业链里抠出了三个利润密度最高、护城河最深的环节

仓位结构是：海力士正股占一半，康宁和日月光各占四分之一。他把这个组合叫做"捏住AI的蛋蛋"

这样看就懂了川沐不赌谁赢AI战争，直接找谁赢都绕不开的环节

存储、封测、光纤，三条AI基础设施的必经之路，供给全部紧张，没有替���选项

$SK 海力士，核心重仓

川沐判断：海力士Q2利润保守预估比美光大一倍，但市值却基本一样。这个错误定价，在ADR上市后会被美国机构迅速纠正

关键催化剂已经到来， 7月10日海力士将在Nasdaq上市ADR，融资约290亿美元，史上最大ADR发行，超过阿里2014年。所有之前买不到韩国正股的美国机构，届时都可以直接买入

$GLW 康宁，三年后的万亿美元公司

康宁本来是做玻璃的，但它垄断了光纤，而光纤是AI数据中心唯一的物理连接介质。今年它又多了一张王牌：玻璃基板

川沐原话："康宁把光互连直接做到自己精通领域的玻璃里面了，之前其他企业还在ABF板上努力，但下一代芯片基板已经变了，降维打击

$ASX 日月光，1000亿市值的小号台积��

所有AI芯片，不管谁设计的，从晶圆出来后都必须经过封装和测试才能变成可用的产品。日月光是全球最大的独立封测厂，台积电CoWoS产能满足不了需求，外溢最大的受益者就是它

川沐逻辑：日月光和康宁的营收利润增速相近，但日月光市值只有康宁的一半。他想用时间来验证，在基本面相近的情况下，市场最终会给哪个更高的定价

[安东尼Anthony @1kbxx]

经过一系列研究，震荡，建仓，集中，我的股票仓位在以下两个主线以及两个副线，直接说仓位，不多说原因了，原因打打太累：

主线1：半导体ETF($SMH)+存储($DRAM+海力士)
这条线无需多言了，仓位最重，全世界的ai主线最核心位置。
A股选择了澜起

主线2：先进封装+玻璃基板
$INTC $TSM $BESI $GLW
A股选择了长电+通富+京东方+拓荆

副线1：全民炒股
$HOOD，A股选择中信证券

副线2：机器人
$BB，A股选择绿的谐波

另外A股有个国产替代线很重要，卖飞了，跌了再考虑建仓吧。

[Bit & Byte @giga_bit_]

MRVL 还是 GLW

[川沐｜Trumoo🐮 @xiaomustock]

下一代AI芯片用玻璃基板代替pcb，康宁是专业的，
下一代AI芯片pcb要换玻璃基板，康宁是专业的，
下一代电互连要换光互连，康宁是专业的，
康宁 $GLW 在玻璃这一块有无与伦比的话语权。

[Bit & Byte @giga_bit_]

$MRVL is clearly winning

[川沐｜Trumoo🐮 @xiaomustock]

如果从十年后看今天的美股，哪一家公司你觉得会成长为英伟达这样的巨无霸.
xai给出了一家公司：IONQ
gpt给出了一家公司：MRVL
老表们以为会是哪一家？

[Rocky - The Stock Trader Hub @RockyTSTH]

$MRVL call buyers are all over the tape 👀🔥

Flow sentiment: Bullish 🟢
Call flow: $12.9M
Put flow: $0
Put/Call: 0.00

100% of premium is hitting calls today.

Biggest focus is the 8/21 $260C, with multiple aggressive buys including a $4.9M sweep above ask and another $1.6M block 📈

Also seeing longer-dated positioning in the 1/21/28 $480C with $2.9M+ traded.

No put flow showing — buyers are clearly leaning upside on $MRVL 🚀

[Marvell Technology @MarvellTech]

Server-grade DDR5 memory costs $27 to $37 per GB today, and prices have risen 300 to 400% since mid-2025. For AI workloads that are consuming memory faster than fabs can produce it, that is a significant infrastructure cost challenge.

Marvell Structera CXL devices address this through a purpose-built hardware block that compresses data at full memory bandwidth as it is written to DRAM and decompresses it on read, completely transparent to the host CPU. The result is that the host sees more memory than physically exists on the device, at ratios that can reach 2:1 or higher on real-world data types.

Director of Product Marketing Arifur Rahman details how the technology works and what it means for the economics of CXL memory pools: https://t.co/KlPv37vcY9

[CapitalFlow @CapitalFlowApp]

Someone is making a serious bet on $MRVL. 👀

[Omer Cheema @Omercheema]

According to GoldmanSachs, CPO Switch BoM Revenue Contribution per Unit (Quantum-X Photonics / NVIDIA):
Estimated supplier revenue per $130k switch (based on BoM values):
NVIDIA (Switch ASICs + Optical Engines design): ~$12k (ASICs) + significant share of engines → major portion.

TSMC (Fabrication/Packaging of ASICs & engines): Substantial share of ~$44.4k combined ASIC/engines value.

Corning (FAUs + SM Fiber): ~$3.6k + $12.3k = ~$15.9k.

Lumentum / Coherent / Sumitomo (ELS + CW Lasers): ~$7.2k.

Senko / others (MPO connectors): ~$5.8k.

Additional (Shuffle Box, assembly): Balance to total BoM ~$75.8k.

[BDigs @B__Digs]

The $NBIS thesis is going mainstream...

[GDP @bookwormengr]

The photonics related stock sell off looks ridiculous.

NPO does require:
- Laser source
- FAU
- PIC/EIC all just like CPO!!!!

The main thing that changes from CPO is where PIC/EIC assembly is located - whether on ASIC substrate (CPO) or off-ASIC substrate (NPO).

And did you know NPO needs more DSP logic (which could be absorbed in ASIC) than CPO, as it needs more correction of disturbance caused by long connection? @austinsemis

So in going from CPO to NPO, you basically you are trading packaging complexity of CPO for packaging simplicity of NPO at higher energy expenditure. Laser, FAU, PIC/EIC all still needed.

What am I missing?

@iamfabian @jaygoldberg @vikramskr @zephyr_z9