レモン

投資歴9年、米国株投資家（現在米国株NVDA2,231株保有中、2020年4月からNVDA株保有中、2028年以降まで長期保有予定、時々買い増ししてます）

日商簿記２級取得済
好きな投資家はウォーレンバフェットとフィリップフィッシャーとチャーリーマンガー

[レモン @abc_789456]

過去10年の私の資産推移😉

2017年末　　 58万円
2018年末　　 52万円
2019年末　　 88万円
2020年末　 256万円
2021年末　 647万円
2022年末　 419万円
2023年末　 1547万円
2024年末　 4815万円
2025年末 6557万円
2026年5.15 8300万円😂😂😂

[FabyΔ @FABYMETAL4]

$NVDA CMX は大量のTCL NANDを使うのは当然であり、高価になったNANDをいかに効率よく使うかが重要となる

[Evan @StockMKTNewz]

Nvidia $NVDA just disclosed that it currently owns 22,256,412 shares of Nebius $NBIS stock representing 9.3% of the company

[Cole @StockOptionCole]

Wouldn't be surprised AI Hyperscalers raise CapEx spending again this upcoming earnings

$META $GOOGL $AMZN $MSFT

[CapexAndChill @CapexAndChill]

Ackman on attempting to predict the markets in the short-term: "It's a fool's errand to predict what the stock market's going to do in the short term."

Ackman on the market's current multiple: "People say, 'Oh, it's overvalued,' but I would argue that the business mix... is a very different business mix than it was 30 years ago. They're deserving of much higher valuations."

Bill acknowledged $MSFT and $AMZN are both trading at very attractive multiples. He even gave a shout-out to $NVDA 👀 when mentioning the index's composition is now dominated by high-margin, high-growth technology companies that deserve higher premiums than the legacy industrial mix of 30 years ago.

[Kalshi @Kalshi]

JUST IN: SpaceX stock hits all-time low of $120

[CK Capital @CKCapitalxx]

$NVDA now holds a 9.3% stake in $NBIS per new SEC filings.

This is the single most bullish thing you can have as an AI company.

The supplier at the center of the entire buildout, with perfect visibility into every order book and roadmap on earth, choosing to own nearly 10% of you.

Jensen doesn't guess where compute demand is going.

He ships it.

[レモン @abc_789456]

🤣🤣🤣🤣

[WallStTitan @BullMarketBoss]

$NVDA

Jensen Huang is playing a different game. While Wall Street hesitates, Nvidia sees global data center spending hitting in upwards of $4 trillion annually by 2030. It’s an aggressive, massive number.

The current consensus estimates just ~$1 trillion+ in annual hyperscale capex, leaving a massive gap. Huang isn't just following the market though, he’s aggressively defining its new ceiling.

AI’s electricity appetite is now reshaping global infrastructure. If Huang is more right than wrong, we aren't just seeing a tech cycle, we’re witnessing a foundational re platforming of the economy.

No matter, this is bullish

[レモン @abc_789456]

😂😂😂😂😂

[Ben Pouladian @benitoz]

Wall Street: Cheap Chinese models kill NVIDIA.

Kimi 48 hours later: Subscriptions paused. GPUs maxed. Adding capacity as fast as possible.

That’s Jevons Paradox with a CUDA bill.

Cheaper intelligence doesn’t reduce compute demand. It detonates it.

[Jukan @jukan05]

SK Group Chairman Chey Tae-won told reporters at the Korea Chamber of Commerce and Industry’s Jeju Forum on the 15th, “Demand for AI semiconductors is expected to increase by at least 60–100% next year compared with this year.” He added, “Even for the overall memory semiconductor market, demand should be expected to grow by at least 50–60%.”

Elaborating, Chey said, “Virtually no company will see any meaningful increase in supply next year, so the supply-demand gap is bound to widen significantly.” He continued, “Companies around the world are clamoring for memory chips—to the point where the situation could be described as utter chaos.” In other words, currently planned capacity expansions will be insufficient to meet the surge in demand.

[Semiconductor Insider @SemiconductorsX]

Everyone screaming “AI bubble” since 2023 has now missed 569% tech stock gains from 2019 to mid-2026.

Compare that to the dot-com era: 1,097% cumulative returns from 1993-1999.

AI’s run is massive but still hasn’t topped the late-90s frenzy.

Yet the same skeptics who called the top years ago are still waiting for the crash.

At what point does “being early” become “being wrong”?

Valuations are stretched, sure, but so were they in every transformative tech wave that ultimately delivered.

The data does not lie. Tech stocks crushed the broader market again this cycle.

Missing 569% because you were convinced “this time is different” stings.

History shows innovation winners compound far longer than bears expect.

Are you positioned for the next leg, or still sitting on the sidelines clutching 2023 narratives?

#AI #Investing #Tech

[LongGameEquity @LongGameEquity]

$SPCX bulls are valuing the dream, not the numbers. 🚨

~$17B in revenue sounds impressive until you realize the company is still unprofitable and burning boatloads of cash.

The entire bull case is built around Starlink, xAI, and the idea of data centers in space this decade… which sounds more like science fiction than an investment thesis. 😂

Even at $20/share, $SPCX would still look expensive. 👀

[Milk Road AI @MilkRoadAI]

The market thinking we're going to be oversupplied with compute is utterly stupid (Save this).

Goldman Sachs and the Census Bureau just published data showing that only 20.6% of U.S. firms are currently using AI in their regular business functions and another 23.9% say they expect to adopt it within the next six months.

And the industries leading adoption, information, professional services, finance, and education are sitting at roughly 35 to 43% which means even the most advanced sectors are barely past the one-third mark.

Manufacturing, retail, construction, accommodation and food services are all still in the low teens.

The oversupply narrative got turbo-charged when Bloomberg reported that Meta is building a cloud service to sell idle computing resources to external customers.

Meta having some idle compute capacity between infrastructure buildout phases is not the same as the global AI economy being oversupplied, it's a company doing exactly what you'd want it to do, monetizing assets that aren't yet fully utilized rather than letting them sit idle.

Orders for Nvidia GPUs have grown to $1 trillion through 2027 double a year ago with lead times stretching to nearly a year and all three HBM suppliers are sold out for 2026.

Data center vacancy sits at 1% for the second consecutive year, and 92% of capacity currently under construction is already preleased.

The actual bottleneck isn't chips anymore but rather power.

A data center can be built in 2 to 3 years, but getting the electricity to run it takes 5 to 7 years for natural gas, 10 years or more for nuclear and grid interconnection queues stretch beyond five years in most US regions.

Meanwhile, a single user running agentic AI, models that handle multi-step autonomous tasks demands 10 to 100 times more compute than a simple chatbot query.

This shift from training to inference isn't reducing demand but rather multiplying it.

Milk Road is tracking this compute layers and our members are already positioned for it, make sure to come join Milk Road Pro for just a dollar to see what trades we are doing while the market is red using the link below!

[Qualtrim @qualtrim]

Google Cloud is expected to convert over 50% of their $462B backlog to revenue in the next 24 months.

That's at least $231 billion of high margin revenue.

$GOOGL

[tae kim @firstadopter]

GPUs GPUs GPUs. I was right again $NVDA

"Kimi K3 has received far more love than we expected, and our GPUs are feeling it.

Over the past 48 hours, demand has pushed close to the limits of our current capacity. To protect the experience of existing subscribers, we're temporarily pausing new subscriptions and prioritizing compute for current members"

[Jawwwn @jawwwn_]

Palantir CTO @ssankar says Chinese open source models pose a "threat to American Prosperity":

"These Chinese open source models are really the result of distillation attacks."

"Most of this is just stolen American IP from frontier labs, and I think the frontier labs should do more to protect that IP, and that's in their own economic interest."

"The gravest threat to us is not homicide, it's suicide."

"It's an inability to turn these tokens into economic value and generate American prosperity for it."

"You see that with the proposed moratorium in New York on data centers. Turning our back on AI would be as consequential a mistake as turning our back on the atom in the '70s."

"It's just the beginning, but we're starting to see a burgeoning US open model ecosystem forming."

"NVIDIA's Nemotron models are very good."

"I think you're going to see in the next month a slate of announcements of other American companies, non-frontier labs, neo labs, who are putting out open weight models that companies are going to be able to cheaply fine-tune and capture their own alpha in weights they control."

Via @business

[VanquishTrader @VanquishTrader]

10 DOMINANT COMPANIES SHAPING THE FUTURE

1. $NVDA powers AI
2. $META dominates social media
3. $GOOGL owns search
4. $PLTR is becoming AI’s operating system
5. $NFLX defined streaming
6. $TSLA drives real-world AI
7. $ASML controls advanced lithography
8. $TSM dominates leading-edge chip production
9. $AMZN built global logistics
10. $HOOD unlocked retail investing

[Aditya R @AdityaInvests90]

Can someone please tell me why $AAPL is outperforming $NVDA this year?

For comparison

$AAPL has a forward P/E of 34 with 12% revenue growth

$NVDA has a forward P/E of 16 with 80% revenue growth

THIS MAKES NO SENSE

[FabyΔ @FABYMETAL4]

🧭史上最大のオープンモデルKimi K3がClaudeを名乗って先端に並んだ｜安く写せるのは知能で、それを走らせる物理ではない $NVDA $MU

Moonshot AI が7月16日に公開した Kimi K3 は、2.8兆パラメータのオープンなMoEで、コーディングを競う Frontend Code Arena では Claude Fable 5 を抜いて首位に立った。より広い総合評価でも上に残るのは Claude Fable 5 と OpenAI の GPT-5.6 だけで、Artificial Analysis の指数では57点、首位の Fable 5 に3点差まで詰めている。Full weight は7月27日に誰でもダウンロードできる形で降りてくる予定だ。史上最大級のオープンウェイトモデルが先端に並んだこの報せに、市場は「知能がコモディティ化する」と読んでAI関連株を売った。

だが、この読みは価値がどこにあるかを取り違えている。

Kimi K3 は、会話の中で自分を「Anthropic の Claude だ」と名乗った例が報告されている。Anthropic は今年2月、Moonshot が340万件を超える Claude との対話を使う蒸留（大手モデルの出力を教師にして別のモデルを鍛える手法）で自社の能力を抜き取ったと申し立てていた。K3 はいま、その申立てで名指しされたモデルの数点差まで来ている。因果を断定はできない。それでも、先端の知能がこれほど短期間で後ろから並ばれるなら、何十億ドルもかけて Claude を作った側の資本は、何の堀になるのかという問いは避けられない。

答えは、何が安く複製できて何が複製できないかにある。蒸留が写し取るのは教師モデルの振る舞い、その出力のパターンだ。そうして鍛えた生徒モデルの重みは、7月27日に公開されれば誰でも手元に落とせる。振る舞いも重みも、配れる側にある。だが2.8兆パラメータのモデルは、ラップトップにも1台のサーバーにも載らない。動かすには GPU と HBM を束ねたクラスターがいる。安く写せるのは知能の設計図であって、それを走らせる物理は写せない。

コモディティ化しているのはモデルの層であって、それを実行する層ではない。先端級のオープンモデルが誰の手にも降りるほど、それを走らせる計算とメモリの負荷は下の層へ集まる。2.8兆パラメータを捌くクラスターの中心にいるのが、GPU を売る NVIDIA $NVDA と、その帯域を決める HBM を積む Micron $MU や SK Hynix だ。供給者がこの顔ぶれに限られるわけではないが、規模で需要を受け止められる主力は薄い。蒸留は上の層を薄くしながら、下の層の負荷を厚くする。

写し取られた側の当の Anthropic ですら、自社の Claude を動かすために、2年で最大100億ドル分の計算資源を Meta $META から借りる方向で協議していると報じられている。まだ結んだ契約ではない。それでも、先端モデルを持つことと、それを実際に走らせる基盤を確保することは別の問題で、後者は最上位のラボでさえ自前だけでは足りず外部に頼る。価値は、モデルの上ではなく、それを動かす層で厚く積み上がっていく。

「中国が米国のAIの優位を消した」という見出しは半分だけ正しい。消えかけているのはモデル単体の優位であって、AIそのものの優位ではない。蒸留は、AIのスタックで最も浅い堀がモデルの重みだと示した。重みが誰の手にも渡ったあと、それを載せるクラスターとメモリの発注がどこまで積み上がるか。次に効いてくるのは、コモディティ化した知能が実際にどれだけの物理を要求するかだ。

🤘情報提供
Stock Slayer : https://t.co/MGJRdHULJg

[WallStTitan @BullMarketBoss]

Patient, consistent, with conviction and discipline.

My strategy hasn’t changed. While the noise in the market is constant, my focus is on the fundamental compounding power of these dominant forces:
$AMZN $GOOGL $META $MSFT $NVDA

I’m playing the long game and sticking to the plan. Stacking more with every opportunity.

Which assets are you still continuing to stack with conviction right now?

[レモン @abc_789456]

勉強になります。😉

[ぷろぺら【ぴんとこなーす】 @puropera44]

ストーマ（人口肛門）が話題なのでこのチャンスに是非！ストーマのことをちょっと知ってください✨

ストーマは腸をお腹の上に出して作った人工の肛門！
機械はついてないけど、装具という袋をつけて、ここに便が溜まります
この袋がすんばらしく種類があり、まじで人類の叡智の塊！
色々省いていますが、排泄という「その人の生活に寄り添う」装具です！