David Timis @DavidTimis

Bernstein has earnings in Memory going parabolic, then collapsing…

Lumentum CEO Michael Hurlston says AI's indium phosphide laser shortage will get worse than the memory crunch, and it's why Nvidia backed both $LITE and its biggest rival, Coherent $COHR. "The numbers of lasers that those kind of customers (telecom) would deploy are in the hundreds, right? Now we're talking about hundreds of millions." "So, for us to get our fabrication facilities up the curve to go from thousands of wafers to millions of wafers is no small feat." "You've covered us for a long time. We've now put online five different wafer fabs. There's a different material. This is not CMOS." "It's called indium phosphide, right? A very complicated name, but it's basically something that can emit light. It's a property, a material that can emit light, and we have five indium phosphide fabs that we're trying to ramp to scale." "We see a huge shortage, and that's one of the reasons, I think, Nvidia invested in Lumentum and invested in one of our largest competitors, who also has incredible indium phosphide manufacturing capability." "Between the two of us, I don't think we can service the demand that Nvidia and others are now putting on us to solve this resistance problem in the data center. And so, the shortage of indium phosphide, I think, will become even more acute than what we see from the memory guys."

Fireside Alpha @firesidealpha

20 hours ago

$LITE's CEO on the shift from copper to optics inside the AI rack: One scale-up rack needs ~8,000 lasers, and he thinks the indium phosphide shortage gets tighter than the one in memory. First deployments in 2H'27. x.com/i/article/2081…

3 14 90 140K 98

"If the multiples on the power, cooling, optical names are correct... Nvidia, memory, they're going up a lot" @GavinSBaker on @theallinpod Since then: Micron near 6x forward, NVIDIA 23x, Vertiv 52x, and memory just sold off again Watch Same rack, different multiples

keyboard_arrow_left Previous keyboard_arrow_right Next

Rare chance to buy Uber below Pelosi & Trump's entry 6 politicians bought $UBER this year and haven't sold: • John Hickenlooper (D) • John Boozman (R) • Donald Trump (R) • Alan Armstrong (R) • Nancy Pelosi (D) • Dan Newhouse (R) Pelosi put up to $1,000,000 on May 29th in Uber call options expiring March 2027 Meanwhile, President Trump put up to $5,000,000 on May 13th And here's where it gets more interesting Uber spent $1,230,000 lobbying Congress in Q2 2026, the same quarter Pelosi bought in

Leopold Aschenbrenner is literally giving you insider trading info He turned $225M into $5.5B in less than 12 months In 2025 he bought: $BE at $18 & is now $297 $LITE at $59 & is now $935 $SNDK at $42 & is now $1,466 Now in 2026, he’s telling you to buy: 1) Applied Digital $APLD 2) Bloom Energy $BE 3)CleanSpark $CLSK 4)CoreWeave $CRWV 5)Intel $INTC 6) IREN $IREN 7) Keel Infrastructure $KEEL 8) Micron $MU 9) Riot $RIOT 10) Sandisk $SNDK 11) T1 Energy $TE 12) Taiwan Semiconductor $TSM Don’t miss out on a generational run

Leopold Aschenbrenner's fund has lost roughly $600,000,000 in a single session today But he's still up over 1,000% on most of these positions from his entry and his puts are printing: • Bloom Energy, $BE, -$210,000,000 (-14.9%) • SanDisk, $SNDK, -$180,000,000 (-11.0%) • Core Scientific, $CORZ, -$54,000,000 (-9.1%) • IREN, $IREN, -$41,000,000 (-8.6%) • Nebius, $NBIS, -$104,000,000 (-4.5%) In 1-3 years, the thesis he's betting on: • AI inference workloads will demand more NAND storage than any prior tech cycle • On-site power generation replaces grid connections for every major data center • Ex-Bitcoin miners become the backbone of AI cloud infrastructure

There is a new bear case on $BE but we disagree (Save this). A recent Invest Like the Best episode argued the US natural gas market could tighten materially starting in 2028. He framed that as bearish for Bloom Energy specifically. Our read is different. The real risk to $BE is that higher prices bring more production online and reduce marginal LNG exports well before physical scarcity actually hits. The real risk is that gas becomes more expensive or harder to secure on a firm basis in the specific regions where Bloom wants to deploy. Our analyst is not selling $BE for three reasons: 1. Efficiency. Bloom's fuel cells run at roughly 54% lifetime electrical efficiency and use less gas per unit of power than most modular engines and simple cycle turbines. As gas gets more valuable, that efficiency edge is worth more, not less. 2. Fuel flexibility. Bloom has longer term paths through renewable natural gas, hydrogen blends, and pure hydrogen. None of those realistically replace conventional gas at gigawatt scale in the near term. 3. Project selection. Bloom can prioritize deployments in regions with strong existing gas infrastructure and customers who can lock in reliable fuel at competitive prices. That directly reduces exposure to the exact regional bottlenecks this thesis is worried about. The stock has already had a rough stretch, down sharply over the past month. But we remain bullish on $BE. Our analysts at Milk Road find underrated gems before the market catches on. We called names like MU, CRDO, NBIS, and BE over the last 3 months. Don't miss the next call, join us for $1 (link in first comment below).

Good expert call on Bloom Energy $BE with a former VP at Plug Power - pretty bullish Hyperscalers did not evaluate Bloom against gas turbines and select Bloom. They selected turbines, discovered they could not get them, and Bloom was the alternative that checked enough boxes. Gas turbines from Mitsubishi, GE Vernova, Siemens and Hitachi remain the incumbent workhorse, but his read is that if the order is not already placed, you are not energizing before 2030. Reciprocating engines sit in the same position: Caterpillar, Jenbacher, Generac, Wärtsilä, all effectively sold out. Transformers, switchgear and substation equipment carry 60 month lead times. What Bloom offered was availability plus modularity. A claimed 90 day time to power on smaller blocks, which he believes is credible at modest scale and unlikely at large scale, plus a build-as-you-go capital profile. Turbines want a single large plant. Behind-the-meter deployment wants building blocks you can add to as long as you have secured the land and the gas tap. > Why the Turbine OEMs Will Not Simply Close the Window Turbine and engine OEMs are deliberately not expanding capacity. They suspect the order book is double and triple booked, and they fear being left with stranded factory capacity when projects fail to reach FID. His analogy is the semiconductor capacity cycle, where consecutive quarters of poor absorption caused structural damage. Their posture, as he characterizes the consensus from trade shows and industry conversation: you cannot buy it from me, you cannot buy it from my competitor, you will wait. If that discipline holds, Bloom's window is measured in years rather than quarters, which is materially longer than the market appears to assume. Bloom's product is closer to a solid state electrochemical device than a precision machined turbine, drawing on an entirely separate supply chain that can be ramped faster. > Levelized Cost: A Premium, But Not a Prohibitive One He built his own LCOE model rather than relying on published work, which he found rested on unexamined assumptions. His output: Gas turbine: roughly 4.5 to 7 cents per kWh Bloom: just over 7 cents unsubsidized, below that with federal incentives Reciprocating gas engine: roughly 8 to 10 cents Diesel: high teens to mid 20s The critical observation is that this is not a 3x premium for speed. That pattern collapses the moment supply normalizes, because buyers drop the expensive option as soon as the cheap one is obtainable. A single digit cent premium does not collapse, because the hyperscaler business case still clears at that price. The offset to Bloom's higher capital cost is efficiency: 60 to 65 percent, against roughly 55 percent for a gas turbine and roughly 45 percent for a reciprocating engine. Bring capex down and the LCOE gap narrows or inverts. > Where Bloom Ranks Today Asked to stack rank for a hyperscaler buyer, he puts Bloom third, behind turbines and engines, purely on track record rather than physics. His analogy: you know exactly what you get from a Caterpillar engine or a GE Vernova turbine the way a Toyota buyer knows what he is getting. No buyer has that reflex for a Bloom box yet. The open questions the buying community has not resolved: real world availability, whether maintenance cadence matches or beats turbine schedules, and the roughly 10 year stack replacement cycle. On that last point he offers a mild positive read-across, noting that in the PEM industry stack rebuild intervals came in longer than originally modeled. The path to second or first place requires two things running together: two to four years of collective industry uptime data, and capex reduction. Oracle, Nebius, Brookfield and AEP are the proof points that will settle it. On whether they will work, he says "the jury is still out," while noting early evidence reads favorably. > Non-Combustion as an Unpriced Permitting Asset The Bloom box does not combust natural gas. It runs an electrochemical reaction. The consequences stack up in a specific and useful way: NOx, SOx and particulate emissions at or very near zero, leaving local air quality unaffected Roughly 65 dBA at three feet, which he compares to a lawnmower at fifty feet, meaning nearby highway noise dominates Zero net water consumption, with startup water recycled as steam Materially easier local permitting Each of those neutralizes a specific community objection, and the pushback is accelerating. New York State's one year moratorium is the marker he points to, alongside complaints in other jurisdictions about power draw, water use and air quality. His honest caveat: to date these attributes have played essentially zero role in purchase decisions. Availability and cost drove everything, and he assumes very little of Bloom's performance so far reflects environmental considerations. If pushback becomes electoral, and he says he is watching whether candidates start running on it, then zero emission on-site generation stops being a nice-to-have and becomes the only permittable option across large parts of the country. He expects this to bite first at the 20, 50 and 100 MW sites going into actual neighborhoods rather than at the West Texas mega-campuses. > Market Share Trajectory Data center demand forecasts he is working from run 40 to 60 GW per year. Bloom's share today sits in single digits. His trajectory: Five years: 15 to 18 percent Ten years: 25 to 28 percent Upside case, if emissions constraints become binding in enough jurisdictions: 40 to 50 percent The constraint that drives the upside case is geographic. Not everyone can replicate what Microsoft and Chevron are doing on the West Texas gas fields. Once data centers have to disperse into places that care about permitting, the zero emissions conversation becomes unavoidable. > The Bear Case He Actually Respects Execution, not demand. He flags this above everything else. Bloom has roughly 1.5 GW deployed against a backlog he characterizes as roughly 20 GW. On Sridhar's own description of the factories, that a visitor will see build activity and factory expansion activity running simultaneously, the expert's reaction is blunt. To an industrial engineer, expanding while still trying to build is a very risky proposition. Doable, but it is the precise point at which fast-scaling companies break, and he notes this is the classic failure mode for startups that find themselves in this position. Q1 was clean. The Q2 print, due around the 28th, is the next checkpoint on whether execution is holding. The secondary risks are demand-side and none of Bloom's own making: hyperscale capex circularity, bubble risk, and whether community pushback genuinely slows the build or simply reroutes it to Texas. > Scandium: Directionally Fair, Materially Overblown On the short thesis that Bloom cannot secure enough scandium, he says the report has some points but overstates them. His rebuttal runs on three tracks. Cost sensitivity. Scandium is a dopant in the zirconium ceramic electrolyte, used at very low concentration, valued because it tolerates the 800 to 900 degree operating temperature. Even if it were 2 percent of materials cost, which he considers extraordinarily high for a dopant, a doubling in price takes it to 4 percent. Bloom likely has the pricing power to pass that through, and a half point efficiency gain would offset it in LCOE terms. His conclusion: more price risk than supply risk over the next couple of years. Supply structure. Scandium is almost never mined primarily. It sits in the tailings of titanium, cobalt, aluminum, iron and lithium operations and is generally left behind. The binding constraint is processing capability, not geological availability, and that processing capacity is being built with national security tailwinds behind it. Scandium-aluminum alloys matter for 3D printing, fighter aircraft skins and missiles, which places it squarely in the critical minerals policy agenda. Company mitigations. Bloom has spent 20 years reducing scandium loading per gigawatt. He located a patent application substituting cerium and yttrium, both more available, and Bloom holds IP on recovering scandium from mine tailings. He reads Bloom's willingness to address the topic directly, rather than deflect, as evidence they take it seriously rather than evidence of vulnerability. Non-Chinese supply exists: he points to Sumitomo's Philippines cobalt operation, which publicly identifies Bloom as a customer. Bloom does not disclose suppliers, and the short report's supply map traces its merchants back toward China. > The Competitive Set FuelCell Energy. Molten carbonate rather than solid oxide, but functionally similar: high temperature, slow start, direct natural gas, suited to stationary baseload. Why they never scaled into this comes down to inertia and strategic drift. Their historical focus was a trigeneration box producing hydrogen, power and heat, deployed for applications like Toyota Mirai fueling at the Port of LA. When hyperscale demand arrived they had nothing to show. His read on the pivot: they saw the multiple Bloom trades at and asked why not us. Ceres Power. UK based, probably second globally in solid oxide IP. Pure licensing model, which means most licensees stay invisible. The disclosed one is Weichai, moving from small C&I units up to hyperscale scale. He doubts Weichai exports into the US successfully but expects success in China. Microturbines and aeroderivatives. TurboCell in the BorgWarner orbit, plus aero engine derivatives repurposed as stationary generators. Everything gets a look right now because buyers are desperate for speed to power. Stealth entrants. He assumes several exist that have not been announced, precisely because Ceres-style licensing deals do not get publicized. Asked whether Bloom owns the US market today, his answer: "Pretty much now they do." > Why Hydrogen Never Worked, and the Read-Through to Plug Useful because he lived it from the inside. Delivered liquid hydrogen bottoms out near $8 per kilogram. Run that through the efficiency stack and fuel cost alone lands around 54 cents per kWh, before equipment, labor, warranty or service. He stopped modeling at that point. Even at a hypothetical $4 per kilogram you land near 25 cents, still a non-starter against a 7 cent Bloom box. Plug built a 3 MW unit at its Latham campus that passed Microsoft's full backup generator protocol, the first non-diesel, non-gas system ever to do so. Microsoft publicized it as a breakthrough and then walked away inside six months once the cost picture clarified. Plug's INVISTA facility was outfitted to build stationary modules for the data center market and effectively none of it shipped. Three sites total, including Calistoga in PG&E territory for public safety shutoff backup, and an EV charging site that existed only because a grid connection was unavailable. Both are showpieces that draw tours. Neither is repeatable. source: Tegus

Memory is the trade everyone loves right now. The largest bitcoin miner's CEO thinks it busts in 18-24 months. Fred Thiel @fgthiel watched the PC-memory boom-bust cycles of the late '80s and '90s up close. His read: Korean makers are enjoying huge paydays today while Chinese capacity gets built fast enough to overshoot demand, same as every prior cycle. Pull up Micron's chart across those decades and the pattern repeats. His one caveat - AI infrastructure dodges that fate longer, because power, not a fab anyone can bring online, is what rations supply for the next four to five years. Jim Chanos @RealJimChanos built a career on exactly this - supply catches up, and the cyclical bulls get caught long at the top. Net for $MU holders: separate the HBM and AI demand from the commodity DRAM cycle. Thiel says the second one still mean-reverts. Thiel spoke with Natalie Brunell @natbrunell on Coin Stories. The full memory-vs-AI-infra split: podcastalpha.substack.com/p/fred-thiel-m…

Podcast Alpha @PodcastAlphaX

a day ago

Fred Thiel @fgthiel runs the largest public bitcoin miner. His number: a mining site costs about $1M per megawatt, all in. An AI site runs $10-15M per megawatt before a single GPU goes in. Same land, same power hookup, 10-15x the build cost. That's why Mara is converting, not

1 3 10 28K 9

Fred Thiel @fgthiel runs the largest public bitcoin miner. His number: a mining site costs about $1M per megawatt, all in. An AI site runs $10-15M per megawatt before a single GPU goes in. Same land, same power hookup, 10-15x the build cost. That's why Mara is converting, not mining. It already controls over 4GW after the Long Ridge and HIF deals, against 1.1GW of live mining today. The catch: Mara had no hyperscaler relationships and no build capability, so it's leaning on Starwood at 80% loan-to-value. The megawatts are real. The tenant demand is borrowed. Chase Lochmiller @ChaseLochmiller of Crusoe makes the same energy-first case - the AI constraint isn't chips, it's power you can actually plug in. The tell for $MARA isn't hash rate anymore. It's how many of those gigawatts turn into signed AI leases. Thiel said it to Natalie Brunell @natbrunell on Coin Stories. How Mara funded the pivot and what breaks it: podcastalpha.substack.com/p/fred-thiel-m…

$MU 5.3x $SKHY 4.1x $SNDK 5.3x The S&P trades near 22x. Semis trade north of 30x. Priced like they're going out of business in two quarters. Ridiculous.

🇺🇸 AI arms race is moving from Silicon Valley to US military bases : - The Pentagon plans to build hyperscale AI data centers across at least 12 US military bases, creating secure compute hubs for next-generation defense AI. - Initial projects at Fort Bliss (Texas) and Dugway Proving Ground (Utah) are expected to attract over $1.3B in private investment with companies like Carlyle, KKR & BlackRock-backed CyrusOne leading development. - These facilities will power AI workloads including battlefield intelligence, autonomous systems, surveillance & Project Maven, while providing the security, land & power infra hyperscalers increasingly need.

keyboard_arrow_left Previous keyboard_arrow_right Next

@Polymarket question is, which hyperscalers will they be powered by?

Memory Liquidation The AI memory market is not the telecom boom, and it is not the housing bubble. What we are seeing now is a leverage event: too much leverage, too much crowding, and too much exposure piled into the same trade, all of which now need to be unwound. That matters because a real supply crunch in memory has been amplified by positioning, acute shortages and sharply higher prices tied to AI infrastructure demand are real. Yes the easy money in the AI trade has been made! Let’s be explicit about what that means. Parabolic charts are not proof of durable fundamentals; they are often evidence of momentum, leverage, and borrowed conviction feeding on themselves. When a trade gets this crowded, price stops reflecting only supply and demand and starts reflecting how much fast money is trapped in the move. The underlying AI demand story is still real, and the fundamental supply-demand imbalance still exists. AI demand has forced companies to fight for dwindling memory supplies, while chipmakers prioritized higher-margin data-center chips and memory prices spiked sharply over the past year. But that does not mean every price swing is fundamental. Narrative follows price: when memory names surge, investors discover scarcity; when they break, they suddenly discover China risk or efficiency gains. That is why the analogies to the 1990s telecom boom and the housing bubble are only partly useful. In those episodes, supply ran ahead of demand, too much fiber, too many houses. Here, demand has outrun supply, but the stock market layered excessive leverage on top of a real bottleneck. As Graham observed, the market is a voting machine in the short term and a weighing machine in the long run. Right now, the vote is being driven by crowding, leverage, and forced selling. Over time, the market will weigh the underlying AI demand and the still-tight supply picture on their merits. What is being liquidated is not the existence of demand. It is the leverage wrapped around the story. Yes Parabolic charts that amplify crowded leverage one way bets should be avoided.

Morgan Stanley: The Paths to 25-50% GenAI ROIC GenAI ROIC Frameworks & Unit Economics Despite surging AI capital expenditures and model training spend, Morgan Stanley is bullish on long-term ROIC, introducing three bottom-up frameworks that point to attractive 25% to 50% ROIC: > Hyperscaler GPU Rental (IaaS): Estimated to generate ~60–70% incremental EBIT margins and 30%+ ROIC. The base-case analysis assumes deployment on NVIDIA GB300 chips with a 75% utilization rate and a rental price of $8.50/hour. > Model-Enabled API (Owned Infrastructure): Estimated to deliver ~70%+ incremental EBIT margins and 40%+ ROIC. Key success drivers include token pricing, token throughput (tokens/second/GPU), and managing the trade-off of dedicating compute capacity toward training versus revenue-generating inference. > Model-Enabled API (Third-Party Infrastructure): Estimated to yield ~30% incremental EBIT margins and ~25% ROIC, accounting for the "middle-man margin" paid for renting third-party compute capacity. Key Structural Trends in GenAI Adoption > Cost Efficiency vs. Revenue Growth: Morgan Stanley’s global AI stock mapping indicates that roughly 80% of near-term AI benefits stem from cost efficiency rather than immediate top-line revenue growth. AI Adopter EBIT margins expanded significantly, doubling the pace of the broader MSCI World index. > Diverging Earnings Revisions: Since late 2023, forward earnings expectations for global companies successfully adopting AI ("AI Adopters") have outpaced disrupted counterparts by roughly 2x, as concrete productivity gains and margin expansions materialize on balance sheets. > The "Enabler" Divergence: In contrast to general corporate adopters, AI Enablers (such as infrastructure providers and data center chip makers) see a heavy tilt toward revenue growth, with roughly 71% deriving major benefits from top-line expansion driven by high-demand hardware and cloud compute sales. $NVDA $AMD $GOOGL $AVGO $AMZN $META $MSFT

keyboard_arrow_left Previous keyboard_arrow_right Next

« On ne fait pas de documentaire sur les gens qui ne gagnent pas. » La bande-annonce est exceptionnelle. Le documentaire s’annonce déjà légendaire. 11 août. Netflix. Mourinho :

Hyperscalers will print cash. Morgan Stanley predicts ~90% net margin in token sales from $NVDA Feynman Data Center. Some hyperscalers like $AMZN and $GOOG use custom ASICs and reportedly achieve even higher inference margins. Couldn’t be more bullish. $AMZN $GOOG $MSFT $META

300 P/E for this btw $TSLA

You aren’t bullish enough on hyperscalers. Morgan Stanley says hyperscalers have 31% ROIC in their AI infrastructure businesses, and 25%-46% ROIC in their inference businesses. Hyperscaler free cash flows will explode if this is true. $MSFT $GOOG $AMZN $ORCL $META

Choose your winner: $GOOG vs. $TSLA #FSD #AV