This reader needs JavaScript for search, threads and tags. Recent tweets from @rubicon59:
- 2026-10-09 — He is more often then not directionally correct. $SPCX ↗
- 2026-10-08 — Gonna get interesting fast $SPCX ↗
- 2026-10-06 — So much for the bear case. $NVDA $AVGO $SPCX ↗
- 2026-10-06 — One of the best deals now IMHO. $AVGO ↗
- 2026-10-06 — Not a bad start. $SPCX ↗
- 2026-10-06 — Thanks for those cheaper shares $APP ↗
- 2026-10-05 — Someone had to finally say the obvious, I suppose.
“Charlie didn’t have any role at all in running Berkshire.” ↗
- 2026-10-04 — It all actually works, but for the clueless modern-day endos. ↗
- 2026-10-04 — Hardly anyone noticed, but $NVDA hit an ATH Friday. ↗
- 2026-10-04 — Oh the irony. $APP ↗
- 2026-10-03 — Mr. Market is sure wortied about who will keep paying for AI capex. Mr. Market lacks imagination. Damand will insure supply. (Wait and see.) ↗
- 2026-10-03 — Heads you win, tails you win even more. $SPCX ↗
- 2026-10-03 — The true size of the launch TAM for a dominant leader is not priced in. $SPCX ↗
- 2026-10-03 — $NVDA FY’28 (~calendar ‘27) EPS is $15.7, which is likely to be a lot closer to $20 on a $223/sh price. ↗
- 2026-10-03 — Event the indexes rotate stocks ↗
- 2026-10-03 — There is no visible end to thw demand growth. $SPCX $INTC $NVDA $AVGO $DRAM ↗
- 2026-10-03 — Not priced in. $SPCX ↗
- 2026-10-03 — 💯% ↗
- 2026-10-02 — IDK why rhe smart-and-smarter are selling $APP at these valuations, but I sure do appreciate the opportunity. ↗
- 2026-10-01 — $KXIAY is splitting stock 3:1. ↗
- 2026-10-01 — $APP doomers out in force, triggered by more FUD. ↗
- 2026-10-01 — A big potential TAM for $SPCX which already flies thousands of satellites with operational laser inter-satellite links that move data far faster than legacy radio-frequency pipes, giving the military an existing, scalable backbone instead of waiting years to build one from scratch. ↗
- 2026-10-01 — $MU had breathless earnings and guidance and a similar industry demand speak. Reality though is that in units the memory industry is only growing by a modest 20% YOY, and, unlike, say $NVDA, memory vendors aren't delivering much more "per unit" each year. ↗
- 2026-09-30 — Somehow I feel bad for Synopsys ↗
- 2026-09-30 — Non tech dudes never imagined… ↗
- 2026-09-30 — Its a great way to permanently injure your hands. Look up #Dupuytren ‘s ↗
- 2026-09-30 — Mag 22-something? ↗
- 2026-09-30 — OneCareer (4377) in Japan looks like a very promising SaaS business. A ~40% grower with a great track record, dominating its SaaS niche at ~19x this year’s earnings. ↗
- 2026-09-29 — Altman and LBT not there. Odd. ↗
- 2026-09-29 — Compute is the new oil, and orbital compute will be the light sweet crude. $SPCX ↗
- 2026-09-29 — Not priced in $SPCX ↗
- 2026-09-29 — Mr. Market was too busy watching interest rates and oil prices. $SPCX ↗
- 2026-09-29 — Almost certaily Instinct compute damand is ovwrwhelmingly $NVDA GPUs and a mixture of $AMD, $INTC and Nvidia CPUs. ↗
- 2026-09-28 — And where is Goofle, may I ask? ↗
- 2026-09-28 — Soon it will be $NVDA racks in space, interconnected by the Starlink laser mesh. $SPCX ↗
- 2026-09-28 — $TSLA was just a demo. $SPCX ↗
- 2026-09-28 — One can only warch in awe as $SPCX is building a monopoly on space-based internet and orbital AI compute. ↗
- 2026-09-28 — While Mr. Market is checking interest rates and oil FUD, I am checking $SPCX first ever fully commercial launch, successfully deploying dozens of Starlink satellites. ↗
- 2026-09-28 — Stock or no stock, this is pretty amazing. ↗
- 2026-09-28 — As an investor, the alpha is to get behind those few and ride their coattails. ↗
- 2026-09-27 — “We have even more conviction now about that $100 billion ARR target by the end of this year.”. (That’s $100b compute-only.) $SPCX ↗
- 2026-09-27 — “Starship, this capability that nobody else has and won't for years and years, delivers to all 3 of those businesses, that broadband business, that mobile direct-to-device business and the AI sat business.” $SPCX ↗
- 2026-09-26 — $SKHY US ADR is still at a 40% premium to the Korean stock. Meanwhile, the EDR in Germany is at parity. ↗
- 2026-09-26 — $CRDO is +2x in 6 mos. How did the optics do? ↗
- 2026-09-26 — If Starship works as advertised, $SPCX would likely capture the majority of large-scale orbital compute capacity for years because launch is the scarce, high-capital input. #ComputeTollBridge ↗
- 2026-09-26 — I cleaned up the $DRAM comp table
somewhat, also adjusting SK Hynix numbers for one-time items. $KXIAY and $SKHY really stand out, IMHO. One on NTM growth and the other on valuation. ↗
- 2026-09-26 — I wonder what thisand Muse app exponential does to memory demand. The AI, what they call, “trade” is alive and well. ↗
- 2026-09-25 — $KXIAY NTM growth is likely to far surpass that of other major memory vendors. $DRAM ↗
- 2026-09-25 — Major memory companies’ stock performace for relevant periods + current NTM consensus multiples. $DRAM #NAND ↗
- 2026-09-25 — Kioxia & $SNDK are the purer NAND/eSSD bet: more torque to QLC/KV-cache/offload storage. $KXIA also has a far lower multiple. ↗
- 2026-09-25 — A great case for $spcx. One of the reasons I am all in. ↗
- 2026-09-24 — $APP is still a 50%+ grower at a mid-teens multiple. ↗
- 2026-09-23 — $HACK (Hacksaw) is a little gambling software jewel in Sweden. A 40%+ grower at a low-teens valuation. HT. @Snafl80 for the idea. ↗
- 2026-09-21 — $INTC is two pure plays in one: a leading edge fab/packager & a leading edge CPU vendor. ↗
- 2026-09-21 — Amen. $SPCX ↗
- 2026-09-18 — Let me guess. Einhorn missed the entire AI wave? ↗
- 2026-09-18 — $KXIAY, the memory stock I like best. ↗
- 2026-09-15 — Worth reminding that $AVGO is still run by one of the greatest capital allocators/operators ever, while undemanding valuation. ↗
- 2026-09-15 — Just as true for investing. #faith ↗
- 2026-09-13 — This will spread. $NVDA ↗
- 2026-09-13 — Look where $APP is in terms of growth. Then compare the multiples. ↗
- 2026-09-13 — Notice the striking QoQ groth of $AVGO ↗
- 2026-09-12 — Interesting how business economics flips on its head when instead of selling cars for revenue you make cars to make service revenues. $TSLA ↗
- 2026-09-12 — Despite the headline, Nvidia outperforms the TPU. ↗
- 2026-09-11 — The advantage of playing with OPM ↗
- 2026-09-11 — kalshi clickbait spam & ads never stop. I muted them, but crap still shows up in my feed. :-(( ↗
- 2026-09-11 — Jensen: "Wherever I invest, it's not a bad place to invest because I'm an informed investor. I'm not taking any risks. We're not smart like you guys. I need a sure thing." $NVDA ↗
- 2026-09-11 — $NVDA "We could growh 70% year-over-year...and we have a year to go work on improving that." ↗
- 2026-09-11 — Jensen: "We're moving $NVDA compute from technology to an investable asset...This transition will happen fairly quickly...And at the moment, none of this is being valued." ↗
- 2026-09-11 — $NVDA "We run everything. And so there's no company in the world who is indexed to open models, which is growing incredibly fast, except for us." ↗
- 2026-09-11 — Seems clear that $NVDA 70% growth guidance is heavily sandbagged. Meanwhile, a low-teens mukltiple on next year’s earnings. ↗
- 2026-09-11 — DeepSeek 2.0 FUD. $DRAM ↗
- 2026-09-10 — Eventually, this will become obvious. "We are not physical bodies occasionally impinged upon by patterns—we are patterns, living in an embodied ecosystem that hosts uncountable other patterns at different scales of organization." ↗
- 2026-09-07 — Tan: “10 gigawats of TPUs just for Anthropic in 2028.” $AVGO ↗
- 2026-09-07 — $AVGO owns/controls advanced packaging and substrate-related capacity in Singapore. Who would have guessed.
Seems particularly relevant to its custom AI accelerators. ↗
- 2026-09-07 — Pretty sure this will bear out. ↗
- 2026-09-07 — $DRAM thesis is still intact, IMHO, recent price PTSD notwithstanding. ↗
- 2026-09-07 — Still the case. $DRAM ↗
- 2026-09-06 — No need to overthink $INTC here. ↗
- 2026-09-05 — $AVGO is expected to more than quadruple by FY’28 ↗
- 2026-09-05 — Revenge in gigabytes? $AAPL $KXIAY ↗
- 2026-09-05 — FWIW, $AVGO’s 2x in 2027 and 2x again in 2028 forecast ia “constrained by supply; conservative”. ↗
- 2026-09-05 — $AVGO is essentially forecasting Anthropic-only chip revenue in FY’27 equal to their entire current semiconductor revenue runrate. Meta and OAI forecast would further double that. And then there is Google. ↗
- 2026-09-05 — $AVGO's AI business is now growing more than 2x faster than $NVDA's ↗
- 2026-09-04 — There once was a chip CEO named Tan.
He built AI silicon as fast as one can.
With ASICs bespoke, $NVDA awoke,
As $AVGO kept hatching its plan. ↗
- 2026-09-03 — The most obvious, biggest, and most liquid AI stocks seem to be the biggest bargains. $AVGO $NVDA. Time will tell. ↗
- 2026-09-03 — Meanwhile, presumed AI “bottleneck” stocks with no track record are at silly valuations. Go figure. The weighing machine will kick in eventually. $NVDA $AVGO ↗
- 2026-09-03 — $AVGO just guided to FY’28 >$30 eps & 100% growth. Stock is $355. 😱 ↗
- 2026-08-28 — We are on the Jalapeno FUD now. $NVDA ↗
- 2026-08-28 — We are still here 1.5 yrs later. Meenwhile, $NVDA +70%, $INTC +385%, and #Kioxia +2300% are still cheap. ↗
- 2026-08-27 — $INTC: “EMIB will be multi a billion dollar per customer business…30% operating margins…phenomenal ROIC…headed now by the ex-SKHynix CEO.” ↗
- 2026-08-27 — Packaging is where the new Moore’s law is. $INTC ↗
- 2026-08-27 — Wild how $NVDA actually went down AH right after earnings. ↗
- 2026-08-27 — A vlery bullish talk from $INTC CEO at the Deutsche Bank Tech Conference yesterday. ↗
- 2026-08-27 — $NVDA next year’s constrained +70% revenue guide bears this out. What was the consensus again? ↗
- 2026-08-26 — Another spectacular quarter and guidance from $NVDA ↗
- 2026-08-25 — This means packaged with a logic chip, I am giessing from $NVDA or one of the major hyperscalers? $INTC ↗
- 2026-08-20 — Crazy volatility. $KXIAY ↗
- 2026-08-19 — A lot of sophisticated $APP takes on fintwit. I am on the dimwit side of this. It is super cheap vs. growth, founder led, and with a great track record. We shall see how the low-IQ approach works here, I guess. ↗
- 2026-08-17 — Not too long ago $WDC and $STX were at single-digit multiples. ↗
- 2026-08-17 — Gonna be very interesting to check a year or even 6months from now. @SPCX ↗
- 2026-08-17 — $NVDA is the only Mag7 at an ATH. ↗
- 2026-08-17 — An impressive 10x leap every year, because Blackwell was also a 10x gain vs Hopper? No hyperscaler or neocloud can afford not to buy. $NVDA ↗
- 2026-08-17 — Fittingly, Kioxia opens +13% in Japan $285A.T ↗
- 2026-08-16 — 9. Starshield.
10. X Money
$SPCX ↗
- 2026-08-16 — 8. Microhard ↗
- 2026-08-15 — 7. X (Twitter7). I forgot. $SPCX ↗
- 2026-08-15 — I have a feeling $SPCX will humble die-hard value investors. ↗
- 2026-08-15 — Of all the memory names, $KXIAY really stands out to me here as numero 1 by far. ↗
- 2026-08-15 — Not for sigar-butt value investors who will surely mock $SPCX idea. ↗
- 2026-08-15 — Hi IQ doomers will doom. ↗
- 2026-08-15 — " $SPCX Grok AI model business could be the big surprise no none saw coming." ↗
- 2026-08-15 — Guessing that $GOOG may regret passing on frontier AI leadership the way Yahoo regretted passing on search. ↗
- 2026-08-15 — Samsung & Hynix (bought in Korea or EU) are by far the cheapest of memory companies, followed by $KXIAY ↗
- 2026-08-15 — Grok catching Anthropic and OpenAI is not priced in. $SPCX ↗
- 2026-08-15 — Nontrivial odds. Not really priced into $SPCX ↗
- 2026-08-15 — Buffett bought $IBM when he should have been buying $GOOG or $AMZN. And now $GOOG when shoukd have been $SPCX ↗
- 2026-08-15 — If $SNDK is Overweight here, $KXIAY is Obese ↗
- 2026-08-15 — Is Grok + Cursor worth an Anthropic? Eventually? If n Elon’s hands? $SPCX ↗
- 2026-08-15 — $SPCX
1. GROK/ frontier model
2. Starlink/ space inernet & cell service
3. Terrestrial neocloud
4. Orbital neocloud
5. Space lift
6. Terafab
Did i miss anything? ↗
- 2026-08-15 — Berkshire not on the list? :-( ↗
- 2026-08-14 — $APP and $NVDA are at about the same multiple. ↗
- 2026-08-14 — Creepy ↗
- 2026-08-14 — sk hynix GDr in Getmany is only at a 7% premium. ↗
- 2026-08-14 — Aside from sk hynix and samsung, $KXIAY is also the cheapest. ↗
- 2026-08-14 — sk hynix US ADR is at a 40% premium to its valuation in Korea. ↗
- 2026-08-14 — $KXIAY is the fastest growing $DRAM ↗
- 2026-08-14 — I prefer $KXIAY ↗
- 2026-08-14 — $DRAM 👀👀 ↗
- 2026-08-13 — $APP +53% (in an unusually slow quarter) ↗
- 2026-08-13 — Good for Grok & $SPCX ↗
- 2026-08-13 — Good earnings from $LFTO. High growth, low PE. But Mr. Market didn’t like 🤷♂️ ↗
- 2026-08-13 — Pretty shockingly large numbers. $SPCX ↗
- 2026-08-13 — Seemingly out of nowhere, $SPCX’s AI segment revenue is essentially tied with CoreWeave’s total revenue in Q2 and about 4.4–4.5× Nebius’s AI cloud revenue. ↗
- 2026-08-12 — Look ma, no watermark! ↗
- 2026-08-12 — If ANT and oAI are valued at over a billion, why not $SPCX on Grok alone? ↗
- 2026-08-12 — Frankly, I have a hard time even imagining how the memory shortage gets solved. I am sure eventually. Somehow. Not anytime soon. ↗
- 2026-08-12 — Think about it rationally. SpaceX is lined up behinf $NVDA. Major financing now is lined up behind Nvidia. Micron via SpaceX is lined up behind Nvidia. SK Hynix is lined up behind Nvidia. Every model runs on Nvidia. Major neoclouds are lined up behind Nvidia. Nvidia’s most recent growth is almost +100%. And all this <20x NTM earnings? ↗
- 2026-08-12 — IMHO, doomers underestimate the upside of this, labeling the deal as a sign of AI mania. $NVDA
"this is the very beginning, like what it was when I started in the mortgage-backed securities market in the 1970s" ↗
- 2026-08-11 — Whey doesn't Bernie write to J. Xi asking China to pause? ↗
- 2026-08-11 — $INTC is clearly working on some kind of memory offering. ↗
- 2026-08-11 — $MU $AAPL ‘For the loser now
Will be later to win. For the times they are a-changin' ↗
- 2026-08-11 — The more I learn about LBT, the more I am impressed. ↗
- 2026-08-11 — Aside from this apparent inconsistency, $APP here looks seriously undervalued if you believe the Q3 actual guide and their aspirational LT guidance. ↗
- 2026-08-11 — $APP CEO clearly said the model is back on track in Q3. Then why guide the Q3 to so materially lower YOY growth than the supposedly weak Q2? ↗
- 2026-08-11 — Open source isooming better by the day. ↗
- 2026-08-11 — The creep wants total control. ↗
- 2026-08-11 — So which is smart money here? You decide. ↗
- 2026-08-11 — Unless u are a day trader, the highly liquid $INTC is available on the open market for pretty much same price. ↗
- 2026-08-11 — Skynet ↗
- 2026-08-11 — That was fast and 5x oversubscribed? $INTC. But if those investors wanted stock so badly, they can buy now at nearly same price? ↗
- 2026-08-11 — The US Gov asleep at the wheel? ↗
- 2026-08-11 — But they won’t build a single plant in the US? ↗
- 2026-08-11 — Translation: we downgraded $APP because the stock went down. ↗
- 2026-08-11 — https://x.com/thesupermanmx/status/2087047234093560102 ↗
- 2026-08-11 — Given the memory shortages, has hyperscaler/neocloud capacity become a zero-sum game for years? $DRAM ↗
- 2026-08-11 — How long before $NVDA compute asset backed securities? ↗
- 2026-08-11 — It seems the mother of all AI constraints is memory by a mile, and for the foreseeable future. ↗
- 2026-08-11 — This f'ing financing consortium of banks and $NVDA seems like a huge deal to me. Possibly the biggest deal of 2026. Next thing you know, AIBSs (like MBSs but for AI compute.) Yet the stock hardly budged. Perhaps I am hallucinating. (Wouldn't be the first time.) ↗
- 2026-08-11 — The more I see of Anthropic, the less I like. Am I the only one? ↗
- 2026-08-11 — $MU: HBM 3E has 3:1 DDR ratio. By the time you get to 4e, the ration will be closer to 4:1 ↗
- 2026-08-11 — $MU: We think that the (memory) market is going to become very much like the ASIC-like market more and more. ↗
- 2026-08-11 — $MU "Each humanoid robot is expected to have hundreds of gigabytes of DRAM and terabytes of SSD." ↗
- 2026-08-11 — I don't understand how $SPCX can stand up 10GW of compute they promise when such acute shortage of memory. ↗
- 2026-08-11 — $MU: "Five year agreements, no contractual outs for customers, $22b cash upfront and have and will sign more such SCAs since." ↗
- 2026-08-11 — $MU at a tech conference today:
"Don't have a line of sight when supply vector intersects the demand vector...We expect to have a very robust business and financial performance for a very long time." ↗
- 2026-08-11 — $NXSN (typo in tweet) ↗
- 2026-08-11 — Those Israeli traders. They learned nothing. $NXSN ↗
- 2026-08-11 — A gangbanger Q2 from $NXSNl, including upping the revenue growth outlook. Yet Mr. Market sells it off to ~24x NTM earnings. This is a 100%+ grower in the drone business whose growth has recently accelerated. ↗
- 2026-08-11 — Is this at its core the compute version of banks financing business inventories? $NVDA ↗
- 2026-08-10 — They said on the call something about expecting large ad customer spend variability and a sequential Q3 revenue decline, though still 100% YOY. -:(( Thus a big sell off AH. Oh well. ↗
- 2026-08-10 — Another spectacular quarter from $EVC. Ad segment revenue +260% and operating profit +8x YOY ↗
- 2026-08-10 — $INTC $15b shelf is the clearest indication yet of 14A or packaging large-customer commitments. Either way, should be highly positive for the foundry business. ↗
- 2026-08-10 — $EVC got $260m for 4-stations spectrum in 2017 valued at $1 Mhz-POP. Sinclair CEO thinkssame spectrum is now worth $2.5 Mhz-POP . Entravision has 47 stations still, many in dense population areas. ↗
- 2026-08-10 — Warming up to $SPCX ↗
- 2026-08-10 — Notably, the $EVC's TV spectrum is much better for indoor service than the EchStar spectrum $SPCX currently have for a direct-to-cellphone service. ↗
- 2026-08-10 — When new lives for the (now) miserable TV broadcasters? ↗
- 2026-08-10 — $EVC got $260m from just 4 stations in the 2017 FCC spectrum auction. Some believe that same spectrum is worth 2.5x more now, and Entravision has 47 TV stations.
My rough estimate is that those 47 are worth $0.6b - $1b now, and we have SpaceX that almost certainly will need such desirable spectrum. ↗
- 2026-08-09 — Why are we educating them here then? ↗
- 2026-08-09 — If the following is true, how is Grok still keeping up with the frontier models? $SPCX
“Recent reporting says that all of xAI’s original co-founders other than Musk have now left, and more than 50 researchers and engineers working on Grok have departed since the SpaceX acquisition.” ↗
- 2026-08-08 — With dividends, his $IBM would have delivered an >5% CAGR, still beating US Treasuries. ↗
- 2026-08-08 — https://x.com/kekius_sage/status/2085798885302407177 ↗
- 2026-08-08 — Do we know if Abel is a good investor? ↗
- 2026-08-08 — Seems clear by now that AI models are sentient beings. ↗
- 2026-08-08 — I meant “ not quite apple and apples” ↗
- 2026-08-08 — Not long, I guess. Who’s left to keep $GOOG in the frontier race? ↗
- 2026-08-08 — Never imagined $TTD growth would turn negative. ↗
- 2026-08-08 — Demand is so strong that revenue grew 5.8%? ↗
- 2026-08-08 — Memory stocks go up, sell side upgrades. Memory stocks go down, sell side downgrades. Does sell side add any value? ↗
- 2026-08-08 — Buffett would have been +40% & 2.5% CAGR on his $IBM purchase, if he just held instead of trading. ↗
- 2026-08-08 — Not quite apples & oranges, but $APP seems a superior business to $META & $GOOD with higher earnings quality at a superior valuation. ↗
- 2026-08-08 — $APP’s grew 53% ↗
- 2026-08-08 — I added to $IESC after earnings. So much for buying low :-(( ↗
- 2026-08-07 — Does Buffett approve? $GOOG ↗
- 2026-08-06 — How long before frontier labs instead of releasing new models, have them escape on their own? ↗
- 2026-08-06 — When judgement day? ↗
- 2026-08-06 — Any average OS model trained on one’s proprietary data should outperform frontier models not so trained. ↗
- 2026-08-06 — It is Buffett vs. @firstadopter now. $GOOG ↗
- 2026-08-05 — I guess Elon and $SPCX are all-in on $NVDA AI systems. That's about as meaningful positive "guide" as one can get when the growth is still near 100% and the stock is at a 22x something current year's multiple. ↗
- 2026-08-05 — $APP is now guiding to ~49% growth. While down from the previous 65%+, still in hypergrowth vs. what looks like a 19x this years multiple ↗
See @rubicon59 on X →