TRUTH-TELLER

Building the free world through cooperation, a U.S.-led new multinational alliance, and coordinated leadership across security, economics, technology

[TRUTH-TELLER @Deilmartin]

@EndicottInvests People who wait for «AI train» will stop are more stupid than thinking that during the Cold War and the Third Industrial Revolution, the USA would have stopped all intelligence, weapon, and industrial production.

[JUST KAWS @JUST_KAWS]

Bank of America added $MU as one of the “best investment ideas”

Micron is trading at a forward p/e of 6

PEG at 0.04

Memory demand is going to last for another 3/4 years

Why are people not buying $MU on every possible dip?!

[Renaissance Europe Institute @reieurope]

(Greek obstruction 🇪🇺🇬🇷)

Greece blocks Russia sanctions after Europe saved it from collapse.

🔸 €256.6bn in assistance since 2010, the largest bailout in euro-area history

🔸 Around 69% of all bailout funds disbursed through the EFSF and ESM went to Greece

🔸 Around €12bn saved annually through exceptionally favourable loan terms

🔸 Greek bonds were included in ECB pandemic purchases despite lacking investment-grade status

Now Athens is obstructing the EU’s 21st sanctions package to protect Greek shipping interests linked to Russian LNG.

https://t.co/MLGhYKdzzK

[Nate Endicott @EndicottInvests]

$NBIS Current analyst price targets:

Northland Securities: $410
Arete Research: $380
Citizens JMP: $270
Goldman Sachs: $267
Compass Point: $260
D.A. Davidson: $250
BWS Financial: $200
Citigroup: $287
Morgan Stanley: $144
BofA: $280
Cantor Fitzgerald: $222
Freedom Capital Markets: $200
Exane BNP Paribas: $255
Baird: $250 (Initiated)

Consensus across these analysts: $262

$NBIS is currently $224, average upside of 17%

[The Tech Investor @TheTechInvest]

🚨JUST IN: $IREN PT RAISED TO $90 AT H.С. WAINWRIGHT

[Thomas James Investing @Thomas_james_1]

Nebius $NBIS will hit ATHs again soon.

That’s it, that’s the tweet.

You already missed the April/May lockout rally, don’t miss this one as well.

[Bilaal- BD investing @bdinvestingg]

If you want exposure to AI , Space & POWER then ETFs are the way :

$DRAM – Invests in memory companies. Top holdings: $SKHY $Mu

$DISK – Invests in data storage companies making hard drives, SSDs and storage hardware. Top holdings: $SNDK Kioxia

$LAZR – Invests in laser and optical networking companies powering AI data centers. Top holdings: $LITE Anthropic $AAOI

$VOLT – Invests in companies supplying electricity for AI and electrification. Top holdings: $GEV $POWL $ETN

$NASA – Invests in space companies building rockets, satellites and communications. Top holdings: $RKLB $SPCX

[TRUTH-TELLER @Deilmartin]

@longinvest32 They will lose everything. Unethical idiots and they will not be able to buy the stock they short

[TRUTH-TELLER @Deilmartin]

@Josevizner Sick person and people deport them all

[TRUTH-TELLER @Deilmartin]

@MotivacionesF These are sick criminal people . Never accept any Argentinians to your country

[Mike Investing @MrMikeInvesting]

There is one stock you simply buy without hesitation on pullbacks…

That stock is $NBIS.

Here’s why:

- $NVDA owns a 10% stake.
- Jensen Huang is heavily bullish.
- Anthropic deal actively in the works.
- Over $50B worth of (2027-2031) contracted backlog.

The market is still underestimating where $NBIS will be within 12-18 months.

$NBIS will make many generational wealth.

Mark my words on that…

[Hardik Shah @AIStockSavvy]

$SNDK | Bernstein 𝐦𝐚𝐢𝐧𝐭𝐚𝐢𝐧𝐬 𝐁𝐮𝐲 on 𝐒𝐚𝐧𝐃𝐢𝐬𝐤, 𝐦𝐚𝐢𝐧𝐭𝐚𝐢𝐧𝐬 𝐏𝐓 𝐚𝐭 $𝟑,𝟎𝟎𝟎

[BDigs @B__Digs]

How $NBIS Stock Became the Neocloud Standout -- https://t.co/16UaJHjKD2
July 21, 2026

Nebius Group was leading the so-called neoclouds higher on Tuesday. The cloud-computing company was getting a lift from further disclosures about its backing from chip maker Nvidia.
Nebius was up nearly 17% on Tuesday, to around $212.97. That was ahead of gains of 7% and 2% for peers like CoreWeave and IREN, respectively, as investors flocked back to the artificial-intelligence trade.

Nebius was getting an additional boost after Nvidia late on Monday disclosed it holds a 9.3% stake in the company, or around 22.26 million shares.

The figure shouldn't come as much of a surprise. Nvidia had already announced a $2 billion investment in Nebius back in March, adding to an existing investment in the company. Nvidia's holding consists of 1.19 million shares held directly and 21.07 million shares held via pre-funded warrants. Nvidia is restricted from exercising or selling the warrant-backed shares until Sept. 11 according to the filing.

Nvidia's investments in companies that purchase or rent its hardware have raised eyebrows among AI skeptics. However, Barron's has argued that it could turn out to be a smart use of its surplus cash, meaning it is not overly reliant on a few big chip customers and locking in demand for future generations of its AI processors.

Nebius has been the standout neocloud -- a company which provides additional cloud-computing capacity -- this year, with its stock having more than doubled in 2026 coming into Tuesday's session. The largest neocloud company is CoreWeave, which is up just 2% in the same period.

Apart from Nvidia's backing, Nebius investors have also welcomed the fact that in March it struck a cloud-computing deal with Meta Platforms that could be worth up to $27 billion. Meanwhile, Nebius announced last week that it had entered into its first senior secured debt facility for approximately $775 million.

"We view the debt financing as another positive data point for the AI infrastructure financing ecosystem, suggesting continued appetite from credit markets to fund contracted AI infrastructure at attractive rates, " wrote BNP Paribas Equity Research analyst Stefan Slowinski in a research note at the time of the announcement.

[Alpha Owl @AlphaOwlTrading]

$NBIS is absolutely ripping today!! 🥵

Did you buy the generational dip sub $200??

[Leopold Aschenbrenner Stock Tracker @LeoATracker]

This wasn't a small trade.

It became one of the fund's highest-conviction investments.

Nebius Group - $NBIS builds AI cloud infrastructure.

GPU compute.

And data centers.

[PhotonBull @PhotonBull]

Reminder to never bet against Jensen and $NVDA

$NBIS

[TRUTH-TELLER @Deilmartin]

These criminals will be be able to join any international tournament

[The New York Times @nytimes]

From @TheAthleticFC: Argentina disgraced themselves and the World Cup final with their charmless petulance, our columnist writes. It was "a tetchy, petulant and distasteful performance by the now deposed world champions." https://t.co/3X3DTE9PFe

[TRUTH-TELLER @Deilmartin]

@pinetree_cap Unethical short-sellers will loose everything. $NBIS increase an other 35-45% coming days

[Milk Road AI @MilkRoadAI]

Micron is going to $4,000 and Kimi K3 is the exact reason why (Save this).

Moonshot AI just released Kimi K3, the largest opencsource model in the world at 2.8 trillion parameters, with a 1 million token context window and always on reasoning mode.

Here's the simple version of why open-source models end up needing so much more memory.

Closed labs like OpenAI or Anthropic run their models on one giant shared system, where millions of users requests get pooled together and processed efficiently on the same hardware so the memory cost per user gets spread thin.

Open-source models like Kimi K3 don't work that way, since any company can download the model and run their own separate copy of it, which means the same memory hungry setup gets duplicated over and over across thousands of independent deployments instead of being shared once.

On top of that duplication problem, every model has to hold onto everything it has already read or generated in a temporary memory buffer called the KV cache, so it doesn't have to reread the whole conversation from scratch every time it generates a new word.

That buffer grows bigger the longer the conversation gets and the more the model thinks before answering.

A 1 million token context window combined with always-on reasoning means that buffer stays enormous for nearly every request, and quantization, which normally shrinks a model's size, barely touches this cache since it scales with conversation length, not model size.

Stack duplicated deployments on top of oversized memory buffers and you get a much bigger total memory bill than a single closed model serving the same number of users.

One analyst desk pointed out that Kimi K3 still requires roughly 1.4 terabytes of HBM per instance, which is an enormous amount of high bandwidth memory demand for a single model deployment.

That single data point reinforces Micron's long-term outlook, because every company running Kimi K3 at scale needs that much memory just to serve one instance and open source adoption tends to multiply the number of instances running in parallel rather than consolidating usage the way a single closed API does.

This is playing out almost exactly like the DeepSeek R1 moment earlier in the cycle, where a cheap, powerful open source model initially spooked chip and memory stocks, but ultimately drove more total compute and memory demand once people realized cheaper inference just increases usage across the board, a dynamic sometimes called Jevons Paradox.

The pricing data backs up how real this squeeze already is.

South Korea's DRAM export unit price has rocketed from around $10,000 per kilogram for most of the past few years to over $100,000 per kilogram as of July 2026, with the move accelerating even further in just the first twenty days of the month.

I remain bullish on memory stocks and Milk Road Pro members are up massively on these trades, come join Milk Road Pro to get our entire thesis for just a dollar using the link below!

[Polymarket Money @PolymarketMoney]

BREAKING: SanDisk spikes after a senior Morgan Stanley analyst calls the recent selloff in SanDisk and Micron a “strong entry point.”