FleetCentere

@Midnight_Captl @jukan05 @paurooteri I’ve never seen someone care so much about what random people say on this known mess of a social media platform

@jukan05 @paurooteri I have supported Jukan from the beginning

This is honestly ridiculous. When I posted 100 bullish posts on memory, nobody said anything. But the moment I post just three cautious ones, people start saying, “Jukan is bearish on memory. We’re done.” “Jukan was a memory bear all along, huh?” That’s the kind of nonsense I’m hearing. I’m still bullish on memory. I have been since 2024.

@mrags316 @AndrewBrandt @TuddleOnRadio @BuffaloBills Then lobby or move, or I guess complain on twitter

@FleetCentere @AndrewBrandt @TuddleOnRadio @BuffaloBills How about the rest of the state of NY? I'm sure people in West Chester and Esssx counties are so thrilled for this new stadium

Paid for with $850 million from public funding. Socialize cost, privatize profit.

@AndrewBrandt @TuddleOnRadio @BuffaloBills City wants team, city pays for team, owner profits. I want sandwich, I pay for sandwich, owner profits. Welcome to capitalism.

@TuddleOnRadio @BuffaloBills lol.

@10009nyny @JulianJung_ @TonySage7237 @Sustainabledud1 lol

@JulianJung_ @FleetCentere @TonySage7237 @Sustainabledud1 Do you feel dumb after wasting your time on this

Shoutout to $CRML, @TonySage7237 and Malcolm Castle for the most shameful PEA I have seen in my life. Kudos to @Sustainabledud1 for all his commentary. The PEA for Tanbreez outlines a whopping pre-tax NPV of US$2.7-3.4B and an IRR of 180%. WTF! 😂 CRML and the CP/QP have some gall! Here are a bunch of fun things I noticed whilst flicking through the study. 1) Tanbreez only has a S-K 1300 compliant resource, no reserve. Not to matter, Malcolm Castle has created a notional reserve equal to 83.5% of the 45 Mt resource. But what is the point? They go on to create a mine plan that extracts over 150 Mt from the much larger management-referenced 4.7 Bt target resource. 2) The highlighted rows in the mining plan don’t reconcile. How is the duration of Phase 4 over years 14-17 equal to 5 years? How is the duration of Phase 5 equal to 5 years? 3) Apparently, no mining losses or dilution factors are required because of the open-pit mining nature. Tough ask given the average 8m thickness of the kakortokite units. 4) The biggest WTF (x 1000) is the processing. CRML report a planned mechanical separation process that has been tested by unnamed independent consultants. They assume a flat 93% recovery in a mechanical process… they also assume 100% payability for a Eudialyte concentrate that is, as @Sustainabledud1 has pointed out, is 3% TREO with a massive silica gel problem. This is mining FRAUD. 5) They also very significant sales of feldspar and arfvedsonite. CRML actually assume they can sell the arfvedsonite locally. I imagine they would have more luck selling ice to the Greenland eskimos. 6) To get an operation of 12.5 mtpa up in Greenland, total capex of US$150M is estimated. This is insane 😂. Includes a large port too! 7) But given they didn’t bother estimating a shipping cost, maybe they shouldn’t have bothered including a port. Anyway, could've written 100 points here but this belongs in the all time mining hall of shame.

@blueprintsmb22 They IPOd at $68, $143 now is 15% annual return, Heisenberg can’t do math

Massively hot IPO that year though. Made like an immediate $500k on my tiny allocation for my book at the fund I was at and sold immediately. Will be interesting to see is Blackstone is right that this is the year of the IPO x.com/Mr_Derivatives…

@sharptechpod @stratechery Bunny Colvin and Hamsterdam died in the name of the code

Subscribe at sharptech.fm or @stratechery to get full episodes of Sharp China, Stratechery Updates & Interviews, Sharp Tech, Greatest of All Talk, Asianometry, and Dithering.

Tim Cook’s Exit and What Comes Next, A SpaceX Deal with Cursor, Q&A on Vibe Coding, TSMC, WhatsApp sharptech.fm/member/episode…

@daniel_koss @aleabitoreddit @JKeynesAlpha @genZinvest0r New management and cash liquidity problems solved, stock is liquid enough for actual size, just need some sellside coverage and some investor relations like an earnings call or two and should capture more attention

@FleetCentere @aleabitoreddit @JKeynesAlpha @genZinvest0r yes... insane quality customers for such a small company.

Holy shit $INTC crushed it! I genuinely believe $SHMD and $LPKF could 10x as a result. They could be the best levered way to profit from Intel's success and this general supercycle. Now I finally understand why @aleabitoreddit @JKeynesAlpha @genZinvest0r and many others love these names! Finally had time to do proper research and DAMN, the upside is real and does seem incredibly asymmetrical. Sorry for the buzzword, but it's accurate. But first: you can verify EVERYTHING I'm writing below with Gemini (don't use other AIs as they suck at browsing and only using recent and complete data)! Except the 10x claim, that is obviously my speculative interpretation, not an objective fact. Here's why I think they can 10x: The $INTC x $SHMD Smoking Guns (95% Certainty Intel is the client): 1️⃣ In March, Schmid shipped their InfinityLine H+ to a "U.S. giant." Intel is the only company on Earth using that massive 700mm panel format for Glass Substrates! 2️⃣ Timeline: Delivery in March → Intel reports "ahead of schedule" yield in April. Adds up perfectly. Coincidence? 3️⃣ Tech: $SHMD owns the "touchless" wet-process tech Intel needs to keep glass from cracking. The $LPKF Smoking Guns (95% Certainty): 1️⃣ Monopoly: CEO confirmed an 80% win rate with "global majors." 2️⃣ You can’t use Schmid’s etcher without LPKF’s LIDE laser. They are technically "married" at the factory level. If Intel is buying the "paving machine" (Schmid), they already bought the "GPS" (LPKF). Why is Intel crushing it? We’re moving from training to execution. Agents (mind-blowing levels of demand and compute) need "Orchestration Compute." Reports show agentic workloads needing (up to) 2 CPUs for every 1 GPU. That's why all CPU stocks are exploding. It has been said for months, but the market is just getting confirmation left and right that it's happening and demand is insane. So insane that companies like Intel are selling inventory they planned to write off! So why can these 2 stocks 10x? Intel is a $300B+ giant. $SHMD and $LPKF are tiny German sub ~$400M nano-caps. A single Intel 'Terafab' order is a rounding error for $INTC, but it represents a massive 15–20% of annual revenue for $LPKF and a staggering 20–30% for $SHMD. What Intel spends on a single pilot line effectively funds these companies for an entire quarter. This is obviously not financial advice. Please do your own research and validate everything in my post with Gemini, seriously do it! That will only harden your conviction and you need conviction to hold these stocks. Even if they 10x their moves will not be a straight line up but insanely volatile! You NEED conviction and I can't give you that, you must earn it. Also, these are nano caps, please don't go full retard and position size accordingly if something goes seriously wrong, which could always happen. They KEY in asymmetric investing is position sizing. If your winners do a 2-10x and some go to 0, you're still up big time. But if you win 9/10 times and then you go to 0 with 1 "all-in swing"... you're dead. P.S. here's the craziest part: while @aleabitoreddit made people aware of $LPFK and it started running, $SHMD is literally trading below it 50 day simple moving average. Their balance sheet is ugly, yes, but if $INTC is the customer, that risk is ridiculously tiny compared to the future upside.

@magicsilicon @dnystedt @intel It’s crazy how his chart in 1965 could predict the B200 name

#OTD April 19, 1965. Gordon Moore published an essay “Cramming more components onto integrated circuits” No exponential is forever, but “forever” can be delayed! 61 years later, we are still cramming more!

@insane_analyst @jukan05 @TBU12345678 You know what Google thinks is non-core, semiconductors

@jukan05 @TBU12345678 I wana know what you two smoking.

If Google were to spin off its TPU division into a separate entity, could it IPO the TPU business independently? Similar to what Chinese Big Tech companies have done?

@TBU12345678 Literally no one is overweight chegg

my favorite thing this season is looking at what sellside is saying about these obvious AI losers that whiff: “Temporary Headwinds Bc AI is Killing $U / $DUOL / $CHGG / whatever, but Thesis Remains Intact, Remain OW”

@ContrarianCurse @benthompson Yea agree that was a dumb take, think he even mentioned recently he was off

@FleetCentere @benthompson When it came out, he was over the moon for that slop

Love and respect @benthompson but good lord and Christ was the Vibes take horrible. No one gives a shit about that app at all

@Midnight_Captl Yes and @stratechery wrote on it too, I take with a grain of salt what CC says outside of where they are directly a counterparty because the ones in the power world all still emphasize the shortage (yea that’s them talking their book too)

@FleetCentere Did you listen to the transcript I posted?

$TSM just told us the exact opposite of this btw

@APompliano Probably not going to be last year though

Only a matter of time...

@kakashiii111 @edzitron He needs Arm to fall to ~$20 to be underwater, seems fine

ARM is down 25% since September 30, 2025, and SoftBank’s LTV is approaching its 2021 high. Son is still on the safe side, for now.

@zephyr_z9 What is the crime here?

Nvidia is like Diddy when it comes to committing chart crimes: "Can't stop, won't stop."

@DKThomp Your partner with one of the best episodes, we like simplicity in single number scores and stats but the compression is lossy podcasts.apple.com/us/podcast/the…

Simplification gets a bad rap, and ofc sometimes it’s cheap branding for something incompletely understood, but it’s cool that the smartest people of all time were geniuses of brevity. E=mc^2 from Einstein. Zero-sum from von Neumann. Thinking is many things, but at its highest level, thinking is compression.

@pstAsiatech @jukan05 What

@jukan05 I believe the initial shipments of H200s to Cuma in February are from inventory. So the numbers here don't add up that this is tied to H200s. Blackwell also using HBM3X...

Just in: Samsung and SK Hynix are increasing HBM3E prices by 20%, driven by the approval of NVIDIA’s H200 exports to China.