Akash

@SixSigmaCapital Jeff sun type

$ARM is approaching 10 ATR's above the SMA 50... Dont shoot the messenger!!

@foliotrail What are the new acquisitions?

$IREN probably approaching 6GW total power with new acquisitions. I underwrite infrastructure assets on contracted cashflows and replacement cost. My model: 2.6GW active AI workloads by 2030 = $13B rev run rate. Market cap today? $19B. The moat is land and power.

@grok this is the entire list 1. Machine Maintenance as a Service 2. Modern Metal Mills 3. Vertically Integrated Defense-Grade Factories You have not heard of 90% of Lockheed Martin's suppliers. You should. Lockheed has more than 12,900 active suppliers across all 50 states and 64 countries. 96% of its supplier footprint is in the US. $6.1 billion a year goes to over 7,500 small businesses. The F-35 alone has more than 2,100 suppliers, nearly half small businesses. Orion has 2,900 suppliers, half small businesses. Not subcontractors in name only. Heart of Georgia Metal Crafters in Eastman, Georgia, precision sheet metal for the C-130, F-16, F-35, and P-3 at the same time. Scotia Technology in Laconia, New Hampshire, titanium and stainless tube assemblies for the F-35 out of 48,000 square feet. Ksaria in Methuen, Massachusetts, fiber optic cables for the F-35 and U-2. One small shop in Cleveland runs the only 50,000-ton forging press on earth and makes F-35 bulkheads out of a single piece of titanium. Every F-35 carries 920 pounds of rare earths. China refines roughly 90% of the global supply. There is a 2027 deadline to eliminate Chinese rare earths from all US weapons systems. 4. Vertically Integrated American Freight 5. An Industrial Lender / Fund for Everything Venture Does Not Touch 6. Vertically Integrated Fabrication (Welding) 7. An American Machine Tool Company With Fast Lead Times In 1981 the US was the world's largest machine tool producer at $5.1 billion. In 2023 the US was fifth at $8.1 billion. China was first at $27.4 billion, producing nearly 2.5x as much as Germany in second place. Japan, Germany, Italy, and the US collectively produce most of the world's machine tools, and the US consumes far more than it produces. US machine tool consumption as a share of GDP is 50% below the global average. China's is 90% above it. Lead times on American and European machines are 4 to 12 months, sometimes longer for 5-axis work. Chinese CNC suppliers are quoting 3 weeks. The precision is catching up fast. The high end (IT5 tolerances, true 5-axis aerospace work) still belongs to Germany, Japan, Switzerland, and a shrinking set of US builders, but the gap is narrowing every year. Machine tools are not a commodity. They are the upstream of everything. If you cannot build the machines that build the parts, you cannot reindustrialize. The company I would build is a focused American machine tool OEM that does three things differently: Ships in under 12 weeks instead of 12 months. Has stock machines for next day delivery. Designs for modular manufacturing, so 80% of the BOM is shared across the product line. Ships with a first-class software and control stack from day one, not a bolted-on HMI from 1998. Start with one product. A high-utilization 3-axis or 4-axis vertical machining center aimed at job shops. Prove the lead time. Expand from there. Or do a saw! For someone I know...

Sure, happy to score each item 0-100 on long-term business viability (market size, scalability, barriers, tech trends, policy tailwinds, etc.). But I can't access the full article details from the linked post. Paste the list here and I'll break it down one by one with reasoning. What's the top item?

If you want to reindustrialize America, this is an amazing list of ideas to work on.

@SunnySaum @CKCapitalxx 😂

@CKCapitalxx Thank you so much for this analysis. It's very clear and to the point. I'll open a short position shortly.

$ONDS holders are about to discover $AVEX and the rotation is going to be violent. Let me show you why. $ONDS trades at 92x trailing revenue and 24x forward. For what exactly? A drone company with limited proven operations, real dilution risk, and half the manufacturing capacity of its peers. The stock runs on hype and retail momentum. The fundamentals have never matched the multiple. $AVEX just IPO’d and here is what you actually get for 4.5x forward sales. 10,000 combat proven drones delivered to warfighters in Ukraine. Two sole-sourced Pentagon contracts worth over $1.2 billion combined. The Army picked AEVEX’s Atlas drone over Raytheon and Anduril. $800 million in 2026 revenue guidance. $503 million funded backlog converting at 96.6%. The revenue is not a projection. It is already contracted. Now let’s put them side by side. Revenue growth: $AVEX guides 85% in 2026. $ONDS does not come close. Manufacturing: $AVEX produces 10 drones per month per thousand square feet. $ONDS produces a fraction of that. $AVEX generates $524,000 in monthly revenue per thousand square feet. $ONDS generates $247,000. Half the output at five times the multiple. Contracts: $AVEX has sole-sourced Pentagon awards and a 96.6% backlog conversion rate. $ONDS has government contracts too but nothing close to $1.2 billion in proven delivery. Dilution: $ONDS has been diluting shareholders relentlessly. $AVEX has $375 million in fresh credit facilities at manageable rates and a thin float that amplifies upside not downside. Valuation: $ONDS at 24x forward. $AVEX at 4.5x forward. AEVEX grows faster, delivers more, has proven manufacturing, and trades at one fifth the multiple. Here is the catalyst that triggers the rotation. The quiet period on $AVEX ends in mid-May. Street coverage initiates. Every analyst that builds a growth-adjusted model puts $AVEX above $ONDS, above AeroVironment, above Kratos. Institutional buyers who cannot justify $ONDS at 24x forward but can justify $AVEX at 4.5x start moving. The thin float means small demand creates large price moves. $ONDS holders chasing drone exposure at 24x forward when 4.5x forward drone exposure with ten times the operational track record just listed on the NYSE. The rotation has not happened yet. That is the opportunity.

@ilzmcfly @Ferrari1735 @MB_Hogan That I do see

It's fine whether the AI is hallucinating or not. But here is the Grok chat confirming if they did or did not deliver explicility from the New Jersey Facility. And there is NONE. All nebius confirmed is that they delivered first tranche. Now according to the air permit and satellite images we can estimate that they are not operational. x.com/i/grok/share/b…

@Ferrari1735 @MB_Hogan @ilzmcfly Just saw too and not able to find

@MB_Hogan @ilzmcfly Hey Mark, couldn't find such quote (the screenshot) in fiscal.ai/seekigalpha transacripts btw, do you know more?

The "Permit Wall": Why Nebius’s $17B Microsoft Deal is Stalled in Jersey 1/ A lot of noise lately about $NBIS delivering its "first tranche" to Microsoft. But look closer at the New Jersey (Vineland) buildout. Despite the hype, the flagship US site is currently a 300MW collection of expensive paperweights. Here’s why. 2/ The Neel Khokhani post is the smoking gun. DataOne (Nebius’s partner) filed for Air Permit PCP250002 in Dec 2025. As of today, April 19, 2026, it is still PENDING. 🛑 No permit = No power. 3/ You might ask: "Didn't they deliver 50MW to Microsoft already?" Yes, but NOT from New Jersey. Arakady confirmed From the March 4, 2026, Morgan Stanley TMT Conference a update that they pivoted to their Mäntsälä, Finland site (expanded to 75MW) to meet the first milestone. New Jersey was supposed to be the source, but it wasn't ready. Arkady: "It allowed us to deliver the immediate compute requirements while we finalize the larger-scale deployments in North America." 4/ Why can't they just turn the GPUs on? 🔌 The Vineland grid can’t handle a 300MW load. The site relies on "behind-the-meter" Mainspring natural gas generators. Under NJ law (N.J.A.C. 7:27-8), you cannot legally operate these without an approved Air Permit. 5/ The NJ DEP isn’t budging. They’ve issued at least two Technical Deficiency notices since Feb 2026. They are questioning emissions data and the "single source" impact on the local environment. This isn’t a rubber-stamp process. 6/ Add to that: Civil Opposition. 📢 Residents 1+ mile away are already filing noise complaints from construction and unauthorized testing. In NJ, "noise" and "air" are linked. Public protests in late March have put the DEP under a microscope. They won't rush an approval now. 7/ What this means for $NBIS: Revenue: Full annual run rate from Microsoft is now pushed to 2027. Delays: If the DEP rules them as a single source the projectwould likely blow past the "Minor Source" limits and trigger Title V (Major Source) requirements. A Title V permit can take 12–24 months longer to approve and requires much stricter federal oversight. 8/ The Bottom Line: Nebius is "shell-complete" in Jersey, but "power-dead." Until that PCP250002 permit moves to "Approved," the GPUs stay dark and the $17B deal stays on a slow-drip from Europe. 📉 9/ Watch the NJ DEP Bulletin on May 6, 2026 for the next update. Until then, any claim that the US flagship is "fully operational" is pure hopium. $NBIS $IREN $CRWV x.com/neel_epochal/s… ft.com/content/f2bae7…

@EndicottInvests I mean how is this even reliable if crwv is at the top

New SemiAnalysis for Neoclouds #1 on their own island is $CRWV #2 is $NBIS along with $ORCL & Azure Not recommended: $IREN

@BlackPantherCap Yes

🚨LAST CHANCE! Ladies and gents, this is insane! 😳 I’ve created a free group on X. +500 comments/request! Listen careful here: +120 investors now. We still got room. BUT… I’ve already tried adding EVERYONE who requested access so far, but X has changed group messaging. THEREFORE you may need to update your X app enable the new X Chat system by opening messages and setting your personal 4-digit PIN first. If you want in, comment: YES Once that’s done, I can add you. No paid access. No guru nonsense. No signals. No sales funnel. The goal is simple: A serious space for like-minded investors to discuss markets, share research, break down news, challenge ideas, and learn from each other in real time. What it is: • Research sharing • Smart market discussion • News flow and reactions • Networking with serious people • Different views, same goal: improve What it is not: • Spam & pump • A sales funnel • Buy/sell alerts • Self-promotion • Financial advice What’s expected: • Add value • Be respectful • Think independently • Debate ideas, not people -BP

@glenchernen @NhazeINhaze How is the 7 estimated

Yousaid it but still left a ton of the story on the table. 4,500MW is actually only base starting point so the 90% remaining is only 90% remaining of declared MW. There's at least another 7GW close to being declared from my estimates. Oh, that's if they wanted to declare it. And I'm sure there's at least double that again that is further away but still closer than anybody else. All you, Gotta do is listen and read everything about this company for four years and then it becomes clear. While others were busy, pinning their hopes on building stacks and stacks of bitcoin, these guys spent years securing electrical generation worthy sites, and they did it before anybody else so they got the sites extremely cheap. I don't know how many sites are acres they have but from everything I've ever read. It leads me to believe that it may be the largest electrical generation land package assembly by any entity on the planet. $IREN 💪

People comparing $NBIS to $IREN don’t understand the difference between renting a room and owning the building. NBIS owned capacity: 75MW (Finland) IREN owned capacity: 4,500MW (Texas, Oklahoma, Canada) That’s 60x. Not 2x. Not 5x. Sixty times. IREN is energizing 1,400MW this month. That single site is 18.67x NBIS’s entire owned footprint. MSFT contract generates $1.94B/yr on just 200MW at 85% EBITDA margins. Only 10% of total capacity is contracted. The other 90% is upside. $15.8B market cap. $37B potential EBITDA at full utilization. GPU rental prices surging 40%. Sweetwater live. Earnings May 13. Bears are running out of talking points. Bulls are running out of time to accumulate. $IREN

artemis is reminding me that when humanity is good, it's SO good.

@pepe_maltese Do $NBIS investors not see the power shortage thesis? Or do they just view mw the same regardless of if they’re not energized or energized

This is classic commodity logic.When the world discovers a shortage, the owners of the boring stuff suddenly become strategic assets. This might not seem much, but the industry structure just revealed that the conditions for pricing power just went up. Don't be amazed if the next $IREN deal has better pricing power than the MSFTs

One of the minds that have shapped my investmsnt philosophy is Jim Rogers. If he was now in his prime and investing in AI infra, I bet he would say: "Shortages make fortunes." When everybody wants AI, but the real shortage is transformers, switchgear, and power, the money moves from glamour to the boring owners of scarce infrastructure.

We see our home planet as a whole, lit up in spectacular blues and browns. A green aurora even lights up the atmosphere. That's us, together, watching as our astronauts make their journey to the Moon.

@CKCapitalxx Not bullish on secured power?

Luckily I own $NBIS and not $IREN

@KawzInvests Can you explain $AEHR, I haven’t looked into it much yet, but earnings seem to be declining? Is there a rerating or part of the story I’m not seeing

Building the next-generation optical chips for AI (800G and 1.6T) is incredibly difficult. At these extreme speeds, signals easily distort, and just one bad chip can crash a massive, multi-million-dollar AI cluster. Stop trying to guess which optical manufacturer (like $AAOI or $COHR) will win the final cloud contracts. Instead, buy the companies that make the testing and inspection equipment. These testing companies are "winner-agnostic tollbooths." Because there is zero room for error in AI data centers, every manufacturer is forced to buy this testing gear right now to validate their new production lines. The testing companies get paid first, no matter whose chip ultimately wins. $VIAV & $AEHR: Provide the essential fiber network testers and "burn-in" equipment to ensure these parts actually survive in the real world $KEYS (Keysight): Sells the advanced oscilloscopes engineers absolutely need to test 1.6T signals in the lab. $CAMT (Camtek): Acts as a "tax on chip complexity" by inspecting the microscopic layers of advanced AI chip packaging before they are stacked.

Texas is deploying battery storage so fast that it's almost keeping up with the other 49 states... by itself!

Why does it cost less to build a base on the moon than high speed rail from sf to la

@DollarCostAvg Is this being energized not priced in?

This Recovery climb has been amazing to watch—> and it’s only getting started. Once the $IREN 1400 deal drops, this thing could go straight vertical 🤣📈

@fundmyfund The show more is

I'm 61 years old and resigned from Palantir. My annual income is 8 million US dollars. My March advice: $KOS (Kosmos Energy) — Strong Buy $BW (Babcock & Wilcox Enterprises) — Buy $FSLY (Fastly Inc.) — Strong Buy $MU (Micron Technology) — Strong Buy $LITE (Lumentum Holdings Inc.) — Buy $ONDS (Ondas Holdings Inc.) — Strong Buy $CRDO (Credo Technology Group) — Buy $WDC (Western Digital) — Strong Buy $AXTI (AXT Inc) — Buy $UCTT (Ultra Clean Holdings) — Strong Buy Show more h/t @grok This shit will get like 1K likes maximum because the algo is amazing - ask @elonmusk

@RKLBMan Not enough money 😔

So many amazing companies to invest in...

@Sandeman52 All that matters to be honest

Know what’s crazy? I made a million dollars today and only 2 of my friends and my mom and wife know who Sandeman52 really is. 🫠