$Bloom Energy Corp(BE)$ is becoming increasingly interesting as AI data center expansion runs into a major bottleneck: physical power infrastructure. Off-grid microgrids can help bypass transmission and grid-connection constraints, giving Bloom a large addressable market as data center power demand continues to grow. That helps explain why Bloom’s backlog has surged to roughly $20B. But the bigger opportunity may be what happens after each system is deployed. Bloom attaches a service contract to 100% of its systems, with contracts typically running 10–20 years. That creates a long-duration recurring revenue stream on top of the initial equipment sale. Bloom estimates this contracted service opportunity at roughly $14B. So the thesis isn’t simply abo

$NVIDIA(NVDA)$ ’s open agent ecosystem now spans AI infrastructure, cloud, cybersecurity, networking, enterprise software, energy and financial services. Infrastructure & Cloud$Dell Technologies Inc.(DELL)$ $LENOVO GROUP(00992)$ $DigitalOcean Holdings, Inc.(DOCN)$ $NEBIUS(NBIS)$ $Hewlett Packard Enterprise(HPE)$ $SUPER MICRO COMPUTER INC(SMCI)$ $CoreWeave, Inc.(CRWV)$ $IREN Ltd(IREN)$

$Nokia Oyj(NOK)$ at $10.29 is entering a very interesting stretch. Three catalysts are landing almost back-to-back: 🧠 Oct 20–22 — NVIDIA GTC$NVIDIA(NVDA)$ ’s biggest AI event of the year puts the entire AI infrastructure ecosystem in focus. 💰 Oct 22 — $NOK Earnings The timing couldn't be tighter. And there’s another connection: $NVDA invested $1B in $NOK, taking roughly a 2.9% stake. That gives the AI infrastructure story another layer of relevance as the market heads into GTC week. 📈 Technically, $NOK is getting tight. The level I’m watching is $11. A clean break above it could open the door toward $15+. 🎯 Risk level: $9.50 🎯 Upside target: $15+ The setup is simple: Tight chart → $11 breakout → AI catalys

The potential $11–13B Monzo deal makes more sense when you look at where $Nu Holdings Ltd.(NU)$ is heading. $NU has already shown ambitions beyond the Americas with its Nu Global strategy. Now Monzo could give it a ready-made platform in Europe: 🇬🇧 ~15M retail customers 🇪🇺 UK base + expansion into Ireland and Spain 💰 $NU generates roughly $1B in quarterly net income Instead of building an EU customer base from scratch, Nu could potentially use Monzo as a launchpad and accelerate expansion across Europe. Near term, the math may not look great. Monzo reportedly commands a much higher valuation multiple than $NU, so the deal may be EPS dilutive initially. But the bigger question is what Monzo could become inside Nu’s global platform. And $NU is already

$Meta Platforms, Inc.(META)$ ’s Muse just launched, and demand is already putting pressure on capacity. Here’s the interesting part 👇 ☁️ Every Muse user gets access to a free cloud computer. Each instance reportedly runs with: 🧠 2 vCPUs → $Advanced Micro Devices(AMD)$ 💾 8GB RAM → $Micron Technology(MU)$ $SK hynix(SKHY)$ 💽 100GB SSD → $SanDisk Corp.(SNDK)$ Scale that across millions of users and the infrastructure requirements start adding up fast. And this is just one AI application. The AI race isn’t only about training bigger models anymore. Every successful AI agent or applica

I was quiet last week. I’ve been locked in. The setup I’m watching is simple: months of compression, multiple independent signals, and a potential volatility expansion in Q4. 📈 $SPDR S&P 500 ETF Trust(SPY)$ Zoom out. The daily chart has been grinding through months of chop, consolidation and fakeouts. If positioning gets too defensive, a volatility expansion could catch the sidelines off guard. ₿ $BTC Bitcoin has quietly held most of its gains since mid-August and is back above $85K. A clean breakout could become another major risk-on lever for markets. 🛢️ $Colgate-Palmolive(CL)$ Oil rejected $100 despite the current macro and geopolitical backdrop. If the market has already priced much of the conflict

🧠 Memory demand is growing faster than supply, yet the entire memory complex has been hit hard. That creates an interesting setup: fundamentals remain tight, while valuations have compressed across the group. Citi expects DRAM and NAND supply to remain below demand through 2028. So which one offers the most compelling risk/reward? Here’s the valuation snapshot I’m watching: $Micron Technology(MU)$ — 6x forward P/E$SK hynix(SKHY)$ — 4x forward P/E$SanDisk Corp.(SNDK)$ — 7x forward P/E Kioxia — 4x forward P/E Samsung — 4x forward P/E The market is clearly pricing these businesses as cyclical memory companies. But the current cycle has a different driver: AI infrastru

The guest list at Trump’s Sept. 24 state dinner for Xi Jinping was basically a who's who of American corporate America. More than 30 business leaders were invited, with tech, semiconductors, finance, energy, aerospace and consumer companies all represented. Here are the major publicly traded companies represented: AI & Semiconductors $NVIDIA(NVDA)$ — Jensen Huang$Advanced Micro Devices(AMD)$ — Lisa Su$Micron Technology(MU)$ — Sanjay Mehrotra$Qualcomm(QCOM)$ — Cristiano Amon Big Tech $Apple(AAPL)$ — Tim Cook$Alphabet(GOOGL)$ — Sundar P

🐯 Hey Tigers! Let’s clear up one thing first: $Oracle(ORCL)$ is not bankrupt. The issue here is much more interesting than that. Oracle needed massive amounts of power for a giant AI campus and couldn’t afford to wait years for a traditional grid connection. That’s where $Bloom Energy Corp(BE)$ came in. ⚡ Bloom reportedly got the first system running in just 55 days, beating its own 90-day target by 35 days. Oracle liked what it saw enough to expand the deal to as much as 2.8 GW. But now there’s a problem. 📍 Project Jupiter in New Mexico Oracle filed a force majeure notice because the natural-gas pipeline intended to supply the campus won’t be completed on the original schedule. The pipeline timeline has r

FUD is everywhere on the timeline right now. Some of it is justified. Most of it? Probably noise. But when fear starts creeping higher — especially when it gets close to maximum levels — that’s when I start paying much closer attention. 👀 We’re also heading into an interesting calendar window: Month-end → soonQuarter-end → soon That can create some ugly-looking headlines and price action over the next week or so. So don’t be surprised if the timeline suddenly gets much more bearish. ⚠️ The key is distinguishing real deterioration from news designed to trigger a reaction. Sometimes you can almost feel when a headline is engineered to spook positioning. That kind of forced flush is exactly the type of event I want to watch for. Not because every selloff is a buying opportunity. Because extre

Jensen just crowned $Marvell Technology(MRVL)$ as a potential next trillion-dollar company. 👀 That sounds aggressive. But the underlying thesis is worth understanding. For years, the AI infrastructure story has been dominated by GPUs. $NVIDIA(NVDA)$ → compute But an AI data center needs far more than compute. Thousands of GPUs need to communicate with each other at extremely high speeds. That requires sophisticated networking equipment and connectivity chips to move massive amounts of data across the system. And that’s where $MRVL comes in. ⚡ The part of AI infrastructure hiding behind the GPUs Marvell supplies semiconductor solutions used for networking and connectivity inside AI data centers. The bigger

The $S&P 500(.SPX)$ hasn’t posted a 1% down day in nearly two months. That kind of price stability is worth paying attention to. 👀 One possible explanation is that larger institutions are gradually building exposure. 🏦 Big funds can’t simply dump massive buy orders into the market at once. A large order can move the price against them, so institutional trading desks often break execution into smaller pieces and spread them over time. That kind of steady demand can help explain why pullbacks have remained relatively contained. 💰 And there’s another factor to watch: There is still a meaningful amount of capital sitting on the sidelines. If that money starts moving back into equities while institutional buying continues, the market could have ano

The Neocloud trade is starting to separate into clear tiers. Instead of treating every AI infrastructure name the same, here’s how I’m currently grouping the space: 🥇 Tier 1 $CoreWeave, Inc.(CRWV)$ $NEBIUS(NBIS)$ These are the names at the top of the current setup. 🥈 Tier 2 $Oracle(ORCL)$ $Alphabet(GOOGL)$ Major cloud platforms with significant AI infrastructure exposure. 🥉 Tier 3 $Microsoft(MSFT)$ Still a major player, but sitting in its own tier in this framework. 📊 Tier 4 $Amazon.com(AMZN)$ $Di

🐯 Hey Tigers! The last few days have been fun, but this is exactly where choppy markets can turn traders upside-down. 📈 Two green days can look like a trend change. 📈 Confidence rises. 📈 Position size gets bigger. Then the counter-move hits and suddenly you’re trapped in the chop. I’m not saying this can’t become a full-on change of character. The setup is getting more constructive for bulls. But I still want higher-timeframe confirmation before adding meaningful size. 👀 Here’s what I want to see: • $SPDR S&P 500 ETF Trust(SPY)$ posts a strong weekly close • Crypto continues to hold, giving us a read on speculative strength • Oil $WTI Crude Oil - main 2611(CLmain)$ keeps bleeding lower • Growth nam

18 months ago, the bear case on $Meridian Hldg(MRDN)$ was straightforward: Too much merger debt. That argument is looking very different today. 👀 📉 The balance sheet has changed dramatically: • Key debt down 46% YoY to $34.6M • Net debt down 65% to just $9.4M • Net leverage down to only 0.39x • $17.3M of cash on the balance sheet That's a very different setup from a small-cap company weighed down by debt. 💰 And the bigger point is what the deleveraging enables. A highly leveraged company has limited room to invest. Cash gets consumed by interest and debt repayment, leaving less flexibility to pursue growth. $MRDN is moving in the opposite direction. The company has been reducing debt while continuing to grow revenue and generate profit. That gives

🤖 Last week, everyone thought AI was over. The narrative was everywhere: AI is slowing down. AI capex has peaked. The AI trade is getting crowded. Meanwhile, the industry kept moving. 👀 🚀 OpenAI released GPT-6 Sol & Luna 🧠 Anthropic released Opus 5.5 ⚡ $SpaceX(SPCX)$ released Grok 4.7 🔥 $Meta Platforms, Inc.(META)$ ’s Muse is going viral The AI slowdown narrative didn't even last a week. And this isn't the first time. There will always be another reason to question the AI trade: • DeepSeek • AI capex concerns • Data center cancellations • Chip export controls • Credit risks • Circular financing • Concerns about the pace of AI progress Every few months, a new headline creates another wave of fear. 📉 Bu

🔥 $Meta Platforms, Inc.(META)$ is building across the entire AI stack. 1️⃣ Chips: MTIA 2️⃣ Infrastructure: Meta Compute 3️⃣ Foundation Models: Muse Spark 4️⃣ Applications: Muse + Meta AI This is bigger than simply building better AI models. The AI race is increasingly becoming a race for vertical integration. The companies that control more layers can optimize the system end-to-end, from compute costs and model efficiency to distribution and monetization. $Alphabet(GOOGL)$ $Microsoft(MSFT)$ $Amazon.com(AMZN)$ $SpaceX(SPCX)$ And every layer matters. Owning more of the stack give

🔥 $ARM Holdings(ARM)$ at $275 is getting interesting. SoftBank is reportedly raising its margin loan tied to Arm to $25B, while the chart continues to look increasingly explosive. The technical setup is fairly clean here. $230 is the key risk level, giving the trade a clearly defined invalidation point. The AI/semiconductor theme remains hot, and Arm CEO Rene Haas has also been highly visible in the media, keeping attention firmly on the name. 📈 Swing trade setup: • Entry: $250–$265 • Risk: Daily close below $237 • First target: $400 • Time frame: Multi-week The idea is to let the stock work while keeping the downside clearly defined. Markets are always moving - and sometimes, the best move is knowing what works for you. With Treasury yields, oil

The AI trade is much bigger than $NVIDIA(NVDA)$ . If you look at the entire AI infrastructure stack, the money is flowing through multiple layers — from compute and cloud capacity all the way to networking, power and data centers. Here’s how I’m mapping it: ☁️ AI Clouds$SpaceX(SPCX)$ $NEBIUS(NBIS)$ $CoreWeave, Inc.(CRWV)$ $IREN Ltd(IREN)$ $SharonAI Holdings Inc.(SHAZ)$ This is the layer closest to the actual demand for AI compute. As AI models become larger and inference workloads grow, companies need more GPU capacity and more speciali