Focused on AI power. Views are my own. Not investment advice.
New York, NY
Joined April 2026
- Tweets1.1K
- Following188
- Followers2.5K
- Likes657
Longview Research retweeted
🚨🚨
For $BE - def a clear cut positive , news hitting now
The commonwealth of Virginia released the 2026 Virginia Energy Plan just released
energy.virginia.gov/public/d…
the authors calling out fuel cells and linear generators specifically 12 different pages of this report mention 5 GW of non-combustible generation
Bloom Energy $BE has their footprint all over this for Potentially a new sales channel for Growing Fuel Cell Support
Longview Research retweeted
Elon Musk says the AI boom hits a WALL this year when we run out of power to run the chips.
He says the whole industry is about to learn a hard lesson in hardware.
Chip output is exploding, but electricity is nearly flat everywhere outside China.
So the two lines cross this year.
He says by the end of the year, we will be making more AI chips than we can actually turn on.
Millions of them pile up with nowhere to plug in.
In his words, the bottleneck for the next year is not chips, it is raw power.
Whoever can turn on the most chips fastest wins the race.
In the end, the AI race comes down to one thing:
Raw electricity.
— Elon Musk (.@elonmusk) on Dwarkesh Patel's (.@dwarkesh_sp) podcast
Longview Research retweeted
$FCEL / $BE / $NBIS
FuelCell Energy is getting hyped again on FinTwit as a alternative fuel-cell play for data centers.
Before piling in, read this clear-eyed breakdown of the actual unit economics.
"It seems like fintwit just invests in stories these days. Fuel cells sound amazing. But after years of this exact same quarter, with share count going from 21 million to 54 million in twelve months to fund it, at some point you have to ask if this is a business or just a science project funded by retail."
Worth watching carefully rather than treating it as a narrative trade.
It seems like $FCEL is the new fintwit favorite again so I will say the quiet part out loud as I have looked into it in the past.
I am all for investing in a company before it is profitable. I own plenty of names that are pre GAAP profitability. But you have to be able to see the path. You have to be able to say, if they grow, the math gets better.
This is not that.
This quarter they did 35.5 million in revenue and it cost them 48.5 million to deliver it. Not operating costs. Just the cost to make the thing they sell. You are down 13 million before you pay a single employee.
And it is not a one time issue. Product is 18 million in, 20.2 million out. Generation, the recurring revenue story everyone pitches, is 8.6 million in, 22 million out. You are losing more than two dollars for every dollar of power you sell. How do you scale that. Scaling just makes the hole bigger.
Then there is a 42.5 million impairment on top of it. That is management admitting the assets they built are not worth what they said they were worth.
It seems like fintwit just invests in stories these days. Fuel cells sound amazing. But after years of this exact same quarter, with share count going from 21 million to 54 million in twelve months to fund it, at some point you have to ask if this is a business or just a science project funded by retail.
I am not here to tell you what to do with your money. I am just looking at this filing and I do not see a path to profitability. Not in the long term. Not ever at this unit economics.
Longview Research retweeted
$NBIS Vineland site amendment for $BE solution was approved on Monday.
Yesterday, $BE revealed Bloom Power Connect, a new deployment system that reduces installation time by more than 40%.
I'd bet my house this new integration is being used onsite at Vineland already.
Bloom frames it as a repeatable model for large-scale power installations. With $ORCL Project Jupiter delays, $NBIS's Vineland site will be Bloom's largest deployment for the foreseeable future at 300MW.
Bloom's next largest single-site installation is 80MW at SK Eternix's project in Korea.
Connecting the dots? I, for one, am liking the odds of EOY delivery.
Longview Research retweeted
It seems the short sellers are at it again when it comes to Bloom $BE. If nothing else works, just rehash the arguments Hindenburg made in 2019 with second-hand data.
As for the quality of their research, I'll give one example -
"Scandium is embedded in the fuel-cell stacks, so the faster those stacks wear out, the more scandium Bloom must source over the life of each installation."
97% of the cell by weight is recycled and goes back to the ceramic cell tape casters so Scandium is not an issue for recycling. Only new cells require more Scandium.
It is up to Bloom if it wants to share data to counter all this, but to think infrastructure investors would not do due diligence before investing billions?
Longview Research retweeted
$CEPL. In addition to the valuation being so attractive (see Part Two of my reports), the big question on Capstone comes down to if they win a Data Center deal. To assess this, we ran a six dimension framework (snipped summary chart here) to run through what we believe a Hyperscale or Data Center buyer would assess Capstone on. Capstone passes this buyer test with flying colors that are measured on anchored requirements of deployment, worth a read and some of your time to understand these dimensions. These all point to a greater likelihood of Capstone getting a data center deal at some point. Good things take time, do your own due diligence. And remember that good things take time. NFA. $BE, $GEV, $BW, $BWA, $ENR, $CAT, $CMI, $EROC, $WULF, $CLSK, $IREN, $CIFR, $BWEN, $FLNC, $SMERY, $BIP, $PRIM, $HUT, $BTDR, $MARA, $AEHR, $ASML
Longview Research retweeted
Datacenters have always kept reciprocating engines on site as emergency standby, kicking in when the grid drops and running a handful of hours a year. That role is changing, as recips are increasingly being repurposed to prime power, energizing datacenters around the clock. We estimate that recip OEMs (e.g., Caterpillar, INNIO, Cummins) have been contracted to supply ~1 GW of BTM power this year, and 4+ GW in each of 2027 and 2028. Our SemiAnalysis Energy Model tracks these BTM contracts OEM-by-OEM, project-by-project. Yet, that is dwarfed by the opportunity to come. (1/3)🧵
Longview Research retweeted
AI infrastructure is a team sport.
Industrial Development Funding’s recent announcement with Bloom Energy and Oaktree is a reminder that no single company builds AI infrastructure alone.
From our work with IDF and Nebius to our expanded partnership with Brookfield, each partner plays a distinct role in delivering the power, capital, and execution these projects demand.
In our latest blog, Bloom’s Kevin Passalacqua explores why partnerships have become a defining advantage in building AI infrastructure: bloomenergy.com/blog/buildin…
Longview Research retweeted
$BE, $ORCL, $NBIS. Cracks me up to see X posters worry about Bloom cause of a delay. The deal that got delayed by Oracle is DELAYED not cancelled and it was already understood that it was going to be pushed back a couple weeks ago. All they need to do now is decide on other routes for the gas pipeline, that's it, which exist but just require time for filing/permitting. The Project Jupiter deal is also NOT in the last reported backlog data because it was announced April 27 2026 in the Q2 of this year and they reported their Q1 a few months ago. What's further, Oracle had to pay Bloom to reserve their allocations of power and all the 2.5GW does not ship all in 2027, 2028, or 2029, the ship schedule is moving back but this is NOT PRICED INTO the financials and people thinking that Oracle has an alternative do not understand what is structurally happening in the electricity allocations right now, there is literally no other option for clean power allocations of this scale. One more thing to think about, this might actually be very bullish Bloom because now the ship slots are opening up for other deals suggesting the backlog COULD GO EVEN HIGHER than the original demand curve (if Oracle deliveries move back, near-term slots open up, increasing revenue and demand curve in this time frame). Try not to be such a sucker and try to follow some facts and realities that put Bloom in this position in the first place. Also, since we had Project Jupiter we have had more Nebius deal flow (who just secured MORE financing for MORE capacity), more Brookfield (5x increase versus prior demand announcement now at $25B of demand volumes), and IDF just last week announced. And these are the ones we know about that are in press releases. Bloom's demand and relevance IS ACCELERATING. NFA, good luck, just hope you guys actually understand the facts instead of just the headlines.
Longview Research retweeted
The margin story at $CEPL is quietly exceptional
Gross margin
FY24: 15.7%
FY25: 27.2%
FY26: 32.0%
+1,630 bps in two years — with Q3 FY26 touching 39%.
This isn’t volume leverage. It’s structural. All this at only a ~15% factory utilisation
• Product GM flipped from -4% to +10%
• Parts/service/FPP is now a 64% GM annuity
• Rentals went 19% → 44%
Gross profit went $14M → $34M (2.4x) on just 16% revenue growth. Mix shift toward recurring service is doing the heavy lifting.
What happens when/if they sign a DC deal?
I’m still blown away that they are profitable on a capacity utilisation as low as they are.
Map much leverage to unlock
Longview Research retweeted
$CEPL. If you are starting your research on Capstone Energy+, starts her. This is Part One, there is also a Part Two for the model and scenarios, all still very relevant and current for newer readers. NFA.
Longview Research retweeted
The Bloom $BE short reports are pricing 2026–30 fuel cells on cell specs Bloom first published between 2010 and 2015 — implicitly assuming no technical progress since.
I machine-read Bloom's entire public record: all 681 patents in full text, every SEC filing and exhibit, all 30 earnings calls. The shorts' scandium math (~37–60 tonnes per GW; Hunterbrook 37, Crossroads 60) is built on 25W per cell — a figure from Bloom's 2010 launch — and 100–300µm electrolytes from 2014–15 patents. The 2021–22 tech decks the shorts cite don't update those numbers; they restate them.
Bloom's own filings since then march one direction — thinner and more powerful. Factor the documented trail in (Tons of Scandium per GW):
~37–60 t/GW — the shorts' assumption (2010–15 specs, frozen)
~15–33 t/GW — implied by Bloom's 2021–2025 filings
~6–15 t/GW — the rib-cell architecture patented 2023–24
At 5 GW/yr — the capacity Bloom markets and the Street models — that's the difference between needing ~185–300 tonnes of scandium a year (shorts' math), ~75–165 (current-filings case), or ~30–75 (patented next-gen). For scale: the entire world produced ~80 tonnes in 2025.
(Scope: DEMAND side only. World supply and origin are not considered here.)
The evidence trail, from Bloom's own documents:
Thickness, in Bloom's own filings: 150–300µm exemplified (2014) → "a thin 115µm electrolyte" (2015) → "100 microns or less… 50 to 100 microns" (2021–2025) → rib-supported 30–70µm, patented 2023–24 (US2024/0379983). Half to a tenth of the old material.
Power: Bloom publicly claimed a "+50% increase in power density" for next-gen cells (2020). Run the physics: ~37W/cell at unchanged efficiency — that alone cuts scandium per GW by a third. And even if you freeze wattage at the 2010 number forever, the documented thinning alone cuts scandium per GW by half to two-thirds.
Chemistry is pinned: 9–11 mol% scandia in every filing through 2026. No materials breakthrough assumed, none needed — thickness × electrode gains do all the work.
The honest part: ALL of this is inference — Nobody outside Bloom knows where production sits today.
Longview Research retweeted
Best BTM cheat sheet I've seen (h/t @mosessutton89)
only missing installed cost estimate + lead time
Longview Research retweeted
Google is the confirmed owner behind Project Tembo (formerly Project Jade), Wyoming's largest data center build: 2.7GW scalable to 10GW, 716 acres in Cheyenne's Switchgrass Industrial Park, targeting 2031 completion. Ownership surfaced through a June 30 county filing, not a company announcement, and the prior "bring-your-own-power" framing has softened into a grid-tied structure.
>Capacity/site: 2.7GW initial, up to 10GW scale-out; four data center halls plus office hub; 716 acres, with a further 32.83-acre office support parcel now in permitting (hearing Aug 13, comment window through July 31).
>Power sourcing: Tallgrass Energy blending Black Hills Energy grid capacity, fuel cells, and its own generation at a dedicated substation. Not fully off-grid despite earlier positioning.
>Contractor swap: Crusoe was replaced by Google subsidiary Jupiter Star Holdings LLC as builder/operator last month. Construction continued through the swap, no pause.
>Regional density: Cheyenne now hosts 10 existing data centers, 5 under construction, 9 in advanced discussions, plus Microsoft (3,500-acre expansion tract, pending annexation) and Meta ($800mn, 960 acres, targeting 2027, currently facing wastewater contamination scrutiny).
Oof looks like Crusoe got the boot? "The company is continuing to advance the project and is working directly with the potential large-load customer. While Crusoe is no longer the development partner for this project, the project continues to progress as planned and is tracking for service to begin in early 2028."
Longview Research retweeted
In $BE read through both short reports and FT article. First report was short hit job & garbage. Second short report, Crossroads Capital, more structured and detailed on supply chain issues in China regarding scarcity of critical material Scandium to produce fuel cells. Most folks who know $BE knew about risk regarding Scandium and the supply chain. $BE sources with Sumitomo in Philippines and Rio Tinto, in Canada and not ‘Dependent’ upon China. At same time, supply chain is a risk not fake. Another concern is $BE margins particularly after $GEV turbine backlog is reduced, the reason is customers have been choosing $BE over $GEV due to availability. Competition for data center energy solutions is real and heated, so $BE margins could be threatened. Lastly, valuation of $BE. Valuation is priced to perfection and with sustainable pricing power & growth, likely too optimistic.
Not an expert on Scandium, Competition with $GEV & Siemens is real and doesn’t take a genius to see valuation is very high.
As for my positioning. I was long $BE long stock and then sold in late May plus was short puts at lower strikes every week. One of the difficulties of trading is realizing when a story stock is near the end of a chapter or act. I scaled back last week on short $BE puts. Now I am flat $BE but still positioned yet only 1/4 size as before with short $BE puts $150 & below as short reports did their damage last week.
Anyways, long post but I also post about trading $BE on regular basis. $BE has been excellent alpha in 2026 and will learn in short order if trade is over.
Longview Research retweeted
The behind-the-meter (BTM) market is expected to grow from $108B to $520B.
Everybody knows $BE, but only a handful are familiar with $CEPL.
$CEPL definitely deserves to be on your radar!
Let me introduce $CEPL:
• Perfectly positioned to land a data center deal soon
• Accelerating revenue growth
• Improving margins
• FCF positive
• Positive net income
• Healthy balance sheet
• Attractive valuation
Here's my investment thesis👇
Longview Research retweeted
So many easy holes to address but will only address a few.. $BE
1. Scandium math is weak sauce. Reproducing Hunterbrook's own inputs yields ~25 t Sc₂O₃/GW "sealed" into product (~33–40 t/GW purchased after yield loss and FRU add-ons) but only if per-cell power is frozen at the 2021–2022 figure of 25 W. Bloom's disclosures show ~7.5× cumulative power-density improvement through 2021 and significantly fewer layers since. The entire deficit rests on per-cell power NOT having improved since 2021; t/GW scales inversely with watts/cell. A 2020 Bloom blog states fuel cells fielded since 2016 had "power density five times greater than its first-generation products," with 2020–2021 next-generation cells adding "an additional 50 percent increase in power density". The July 7, 2026 Bloom blog by COO restates that optimizing the doped electrolyte yields "a stack that requires significantly fewer layers of fuel cells to generate the same amount of electrical power."
Sensitivity table below (electrolyte thickness vs. watts per cell): scandium deficit is not fundamentally about geology but rather about an assumed cell-power spec. If that spec is stale, the conclusion weakens materially
2. Supply Elasticity where a Byproduct Market Is Not Fixed
Historical USGS series (Sc₂O₃ equivalent): 2019 ≈ 15–20 t supply/consumption; 2020–2021 ≈ 15–25 t; 2022 production ~40 t; 2024 production ~40 t, consumption 30–40 t, capacity ~80 t (MCS 2025); 2025 production ~80 t, consumption ~60 t, capacity >90 t (MCS 2026). Production doubled 40→80 t and consumption rose ~50% in one year, with capacity already exceeding production. CAGR 2022→2025 ≈ 26%/yr; 2020→2025 ≈ 32%/yr. Hunterbrook's 79→240 t by 2030 implies ~25%/yr a continuation of the observed trend not far fetched at all...
The binding historical constraint has been demand, not supply. With demand now present and a price signal, byproduct capacity is being built. Here's a quick table on ex-China project pipeline...
Hunterbrook adds ~35 t/yr of field-replaceable-unit demand on top of 185 t (new deployments) to reach ~220 t, treating FRU scandium as net-new. But scandium in a replaced stack is recoverable. Bloom operates repair-and-overhaul facilities in Newark, DE (~56,000 sq ft per the 2024 10-K). Patents (US10651479 "Method of recovering metal compounds from solid oxide fuel cell scrap"; US10036082 zirconium/scandium extraction from used SOFCs) and academic work confirm ScSZ recovery is established (acid dissolution + solvent extraction). If Bloom recovers 30–50% of FRU scandium in a closed loop, the ~35 t add-on falls ~11–18 t/yr. The July 7, 2026 blog claims proprietary recovery from waste streams; closed-loop recovery on returned stacks (concentrated, known composition) is materially easier than from dilute ore.
3. Other claims - shows a lack of basic knowledge of project finance
Related-party revenue: this was already flagged by auditor (so not news)... figure is disclosed, audited, GAAP-compliant on transfer of control (not fraud)
Backlog vs. RPO: BLOOM IS NOT A SaaS COMPANY!! RPO is a narrow ASC 606 snapshot that explicitly excludes short-duration and right-to-invoice obligations, and Bloom’s Q1 revenue exceeded beginning RPO by almost 80%, proving RPO is not the order book. Bloom's backlog is LONG DURATION with terms ranging from 5-20 years
- Peer comp is apples to oranges: GE defines backlog=RPO. Software has committed subs / capacity, etc.
-Bloom’s historical system backlog has converted into product / installation / electricity revenue at 40–80% of beginning backlog annually, while service backlog naturally converts at only 2–4% per year because it is a 5–20 year O&M stream. If the system backlog were just not real demand, you would not see 40–80% of beginning system backlog turning into annual revenue
-If RPO were the true order book, Bloom should not be able to generate $751M of quarterly revenue against only $493M of period-end RPO but it did. The reason is accounting mechanics where revenue can be booked, delivered, invoiced, and recognized without ever sitting in period-end RPO.
Summary on points 1-2:
1. Stale input. 25 W/cell is 2021–2022; Bloom has ~7.5× cumulative power-density gains through 2021 and "significantly fewer layers" since. Freezing per-cell power maximizes the deficit and is mislabeled conservative
2. False independence of Methods 1 and 2. Method 1 (patent physics → ~25 t/GW) and Method 2 is a market-share triangulation built from opaque estimates of global scandium demand, SOFC share, Bloom share, and 2025 deployed GW including FRUs. If any of those market inputs are off, the 37–43 t/GW confirmation moves materially. Can be a sanity check, but not an independent proof that Bloom’s current cell spec is still 25 W/cell.
3. Byproduct fixed error. Probability-weighting discrete projects like a fixed mine schedule ignores that byproduct capacity (>90 t vs. 80 t produced) expands with demand and price as it just did (40→80 t in one year).
4. FRU double count. Field replacement scandium is recyclable from the very units replaced; counting it as net-new overstates 2030 needs by up to ~35 t.
5. Single-source supplier quotes. Key claims lean on a Hunan Oriental sales rep in a chat an interested party with incentive to overstate its role. WTF is this institutional grade evidence...shit looks like my uncle's WeChat
NEW: The CEO of an AI giant said again and again: "There is no China supply chain for us."
Trade records, satellites, Chinese filings and even its own suppliers in China say otherwise.
The truth raises an existential threat to Bloom Energy’s story.
Our $BE investigation:
Longview Research retweeted
Congrats on the $BE takedown. As I mentioned, it’s theoretically possible you and others before you with a very similar thesis are ultimately right and Bloom proves to be the Theranos of this AI cycle. I just don’t see anywhere near enough evidence for that today. Every thesis deserves consideration, but it’s equally important to separate facts from assumptions. It would strengthen your case to outline the key blind spots, what would invalidate your thesis, and where you’re inferring conclusions without full visibility into Bloom’s technology roadmap, future materials intensity, and the long-term scandium supply/demand equation.
Longview Research retweeted
$BE and Satish Chitoori (responsible for supply chains), opening up about Scandium Oxide -
bloomenergy.com/blog/demysti…
Longview Research retweeted
If you think BTM power equipment demand or growth rate falls flat then BE will go down, no doubt. And I think your point that mgmt is less than forthcoming about their Sc supply chain is fair/accurate. I've followed and been involved in BE for 15 yrs public and private, and they have enough Sc for their current needs and prob up to 2 GW run rate by most guesses I make. Mgmt will not provide the necessary color. That said, there are *multiple* options to stand up additional Sc supply within 2 years both from new mining projects ($SRL.AX $SCY.V both have ones) and potentially from additional Ti sulfate Sc processing circuits. So while mgmt is not providing full disclosure on their "25 GW/yr" figure (i.e. they CANNOT meet this today, no way),they CAN meet this with additional offtake/supply agreements and those agreements can be providing material within 2 years imo which is under the required timeline. I didn't read your whole report, so if you have other reasons to be short besides this then go for it, but Sc is not a good short thesis (told Hunterbrook same thing). This is a solvable supply chain issue (and more of a reason to be long Scandium suppliers than anything else).