Rob Ptrsn

@mzuhair123 @insane_analyst Thx. In test he mentioned Rigaku via Onto. Have to reread to go over the most crucial.

This is a gem of a talk by @insane_analyst. However, since we have a bad attention span, I have summarized all the key bottlenecks identified by him, just to have a track of what he is looking at. The bottlenecks are sorted by their respective intensity, in descending order: 🔴 Indium phosphide - the worst bottleneck nobody is talking about. CPO laser demand is absolutely murdering supply. The entire chain from substrate to epitaxy to laser printing is a disaster. Key names: Lumentum ($LITE), Coherent ($COHR), Sumitomo, AIXTRON, AXT, IQE, Tower Semi ($TSEM) 🔴 DRAM / HBM - only 3 companies on earth can make it, and they're all sold out. No new capacity coming anytime soon. SK Hynix is leading on HBM4, Samsung is catching up, and Micron is behind. Key names: SK Hynix, Samsung, Micron ($MU) 🟠 Logic fabs - still tight but improving. TSMC doesn't have enough capacity and would rather print N3 at 70% margins than do CoWoS packaging. Intel and Samsung are filling some of the gap. Key names: TSMC, Samsung, Intel ($INTC) 🟠 NAND - high shortage but more prone to oversupply cycles than DRAM. Less attractive as a long-term bet. Key names: Kioxia, SK Hynix, Micron, Western Digital 🟡 High-speed PCBs - the 200G SerDes capable boards are basically sold out across the board. Key names: TTM Technologies 🟡 CPUs - bigger problem than people think. AMD is stuck choosing between GPU and CPU wafer allocation. Only Qualcomm has spare TSMC capacity, but it has failed at data center CPUs 3 times already. Key names: Intel, AMD ($AMD), Qualcomm ($QCOM) 🟢 Power semis - not a bottleneck today but the most exciting upcoming opportunity. Solid state transformers outside data centers are the real play here, not just inside the rack. Watch this space for 2H 2027. Key names: Wolfspeed ($WOLF), Navitas, TI, Infineon, SolarEdge, Enphase, Vertiv, Delta 💡 The wildcard: Semtech. Makes analog amplifiers and equalizers that work across active copper, PCBs, optics, CPO, and more. A near duopoly with MACOM and Semtech parts is simply better. Still massively underappreciated. NFA. Full interview credit: @insane_analyst x @chrisbarber

@ParadisLabs Since seeing PENG last week it’s run like a scalded dog. Not enough days to look into. Now I guess wait for pullback. Reminds me of Druck quote “buy first, investigate after…”

Seems like the two most popular names were: 1. $PENG 2. $OUST I'll start doing research into both. I already have high-level notes for $OUST, but would prefer to beef it out a bit before talking about them. Note: I currently don't have a position in either.

@FinnStockinger Since seeing it it’s run like a scalded dog. Not enough days to look into. Now I guess wait for pullback. Reminds me of Druck quote “buy first, investigate after…”

$PENG: Deciphering the Rally – From the Wild Pre-Market to the Hard Facts from Needham I posted earlier this morning about the insane +20% move in pre-market, wondering if the market had finally "snapped." Currently, with the session in full swing, $PENG is holding a solid +10%, and I’ve finally connected the dots on what triggered this fuse. Yesterday, May 12th, the 21st Annual Needham Technology & Media Conference took place in New York. The new CEO, Kash Shaikh, met with funds and analysts there. The immediate result? Rosenblatt hiked their Price Target (PT) for the company from $32 to $54. Huge shoutout to @ThematicTrader for watching through the entire conference and pulling these critical quotes. Here is why the "smart money" is aggressively moving into positions: 🔹 "There's a lot more demand for CPUs and memory beyond just HBM. Memory is becoming a key requirement in scaling these workloads." 👉 My Take: The market is fixated on HBM, but Penguin is designing the entire memory fabric. They are the systems layer that manages data flow - without them, even the fastest chips are just sitting in a traffic jam. 🔹 "Lots more need for memory in inference to make decisions. Having CXL KV Cache allows the LLM to respond much faster and easier." 👉 My Take: This is the heart of the Agentic AI era. Inference requires minimal latency. Penguin is the only player with a ready-made MemoryAI™ product that solves the KV Cache bottleneck at the hardware level. This is their deepest moat. 🔹 "Large enterprises and governments are using inference and agentic AI on premise. We built one of the largest sovereign AI Factories in South Korea and working with other sovereign clients." 👉 My Take: Sovereign AI is a massive, underrated catalyst. Nations don't want the cloud; they want their own AI factories under full control. Penguin delivers these "turnkey." This is a stable, high-margin business. 🔹 "The AI business is >$1B run rate and growing >50%." 👉 My Take: Find me another company at this market cap with a billion-dollar run rate growing at 50%. This highlights the scale of the $PENG undervaluation compared to the rest of the AI Infrastructure sector. 🔹 "We have a deep relationship with SK Hynix. SK is an investor in the company and based on our relationships, we are able to have access." 👉 My Take: In a world of supply shortages, having an HBM giant as an investor is a massive insurance policy. Penguin gets priority access to components while everyone else is stuck in line. 🔹 "Bookings are a leading indicator and are strong... challenge is time to book revenue is increasing." 👉 My Take: This is a key detail for patient investors. The money is already in the bag (backlog); it’s just waiting for deployment finalization. This is a coiled spring for the FY2027 results. ⬇️Summary: It is incredibly rewarding to see the thesis I've been hammering home finally being validated by both the market price action and institutional analysts. Seeing my deep understanding of the memory bottleneck and the shift toward sovereign infrastructure play out in real-time only reinforces my conviction in the long-term roadmap. This momentum has me more locked in than ever on the vision - I’m fully committed to the Long $PENG. I’m 🐧🐧🐧@WeAre_Penguin Do you see real long-term potential in $PENG, or do you think this is just another hype cycle driven by an analyst report?

@GrowthAdvisorHQ @michaelsikand @joinautopilot @pennycheck Yep that gave me pause. @michaelsikand any insights there?

@michaelsikand @joinautopilot @pennycheck I'm surprised they are only guiding for 12% revenue growth this year

I just bought $2M of a brand new stock after it crashed 7% today. $PENG is now a 20% position in my Asymmetrical Bets fund (+89% YTD) on @joinautopilot followed by $10M. Credit goes to legend @pennycheck for being the first to call this stock. With Penguin Solutions I now own the winner agnostic integrator behind the memory, CPU, and photonics supercycle at under 17x forward earnings. 1) The memory business alone is worth the market cap. Penguin's Integrated Memory biz = they take raw DRAM chips from manufacturers like SK Hynix and package them into custom memory modules built to spec for AI servers, telco gear, and enterprise systems. It's now 50% of revenue, did $172M last quarter, growing 63% YoY, ~$800M annualized. Apply a 3x price to sales on just this unit and you're already above what $PENG is worth today. 2) Play the CPU supercycle. CPU:GPU ratios going from 1:8 to 1:1 as agentic AI takes over. $PENG is the lead integration partner for AMD EPYC and Intel Xeon. Every new socket = more memory cooling and integration revenue baked in. 3) The AI Factory platform is real. OriginAI is their turnkey deployment from 256 to 16,000+ GPU clusters for sovereign and enterprise customers. 85,000 GPUs already deployed. UBS says non hyperscaler buyers (sovereigns, neoclouds, enterprises) capture 48% of AI infra spend in 2026. Hyperscalers build in house. But these other players ALL need Penguin. 4) Photonics is the unpriced asymmetric bet. $PENG called photonics early and was an early investor in Celestial AI. $MRVL acquired it $3.25B in December. Now Penguin is building the Photonic Memory Appliance, making it the only public play on this kind of wild photonics tech. The PMA is basically a box that uses light to link memory across a bunch of servers so the entire AI cluster can share one giant pool of memory like it's one big computer. Marvell guides Celestial to $1B revenue in 2029. If Penguin captures even low double digits of that stream, that could be 9 figs of unpriced networking revenue on $PENG's highest margin, most defensible IP. 5) People/partners are cracked. Chairman of $PENG is ALSO Chairman of $LITE. AMD CTO Mark Papermaster sits on the board SK Telecom dropped $200M as a strategic investor New CPO Ian Colle ran AI infra at AWS 6) Risks are real but manageable Penguin's AI cluster business is lumpy and one big customer slipping a quarter can tank earnings (already happened in Q2, down 42% YoY). The memory shortage is a headwind as high DRAM prices are slowing customer orders and hitting Penguin's gross margins. The photonics upside is a 2027+ story, so if it slips, the stock can sit dead money for a while. Because the multiple is still so cheap, I overall see limited downside compared to the upside if their photonics option can be quantified with $MRVL where I could see Penguin trading closer to a 30x+ forward PE. Surf's up. Full thesis linked on Substack below.

@KairosPraxis It might be the best book you’ll ever read. Hits all geographies, trade, dynasties, Econ, industrialization, geopolitics. Took me years to finish it. Please stick with it and report back

Also weekend reading. Guess I'm a commodities trader now

Weekend reading: Huge deep dive of $HUBN compiled by my research assistant. Let's see if it's any good. Huber-Suhner $HUBN is a 🇨🇭 company benefitting from a lot of AI-related tailwinds. 1. Optical Circuit Switching (OCS) - Uses piezoelectric actuators, different form $LITE's MEMS mirrors. 2. Fiber optics: Chassis, cables, and multiplexers to squeeze more bandwidth. $MSFT partners with them for better scaling across datacenters. Datacenter offerings only contribute 31% of revenue, which seems like an opportunity. They also have some cool defence products (RF Assemblies, fiber optic interconnects), which contributes 37% of revenue.

@PhotonCap So for the specific AI data‑center SOI+SiN stacks, “SiN” is mainly provided by the SiPh foundries themselves as a deposited film on Soitec‑class SOI, while specialized SiN‑PIC vendors like LioniX and LIGENTEC dominate more niche, non‑datacenter SiN platforms

$SOI / $SLOIF Do you know what SOITEC is and what it does? photoncap.net/p/the-wafer-th…

@ThematicTrader another theme that popped up for me was Edge AI compute. own #LTRX. watching OSS

The AI infrastructure expansion is hitting physical walls: Thermal, Yield, and Energy I've put together 5 baskets that pinpoint the companies solving the next generation solutions or bottlenecks of 2026: 1. Advanced Packaging & Hybrid Bonding $KLIC: Dominating TCB assembly for HBM stacks $ADEA: Holding the essential IP for 3D stacking $ASYS: Precision thermal systems for 3D-IC annealing 2. Material & Thermal Bottlenecks $GSM: The primary Western source for Silicon Metal $PDFS: Using AI analytics to solve the Intel 18A yield issue $MTRN: Beryllium monopoly for extreme heat and defense $KULR: Essential fire-blocking for high-density AI clusters $MKSI: Laser precision for CPO and glass alignment 3. Uranium & SMR Boom $ASPI: The Quantum Laser Enrichment (QLE) potential IPO $LEU: The sole U.S. enrichment strategic asset $UUUU: The Western processing hub for Uranium and Rare Earths 4. The InP Alternatives $POET: Slashing CPO costs with the Optical Interposer $ALMU: Leveraging standardized silicon wafers to achieve similar outcomes 5. The Glass Substrate Shift $SHMD: First-mover in specialized glass processing tools $GLW: The king of high-purity glass material substrates $AMAT: The liquid giant scaling glass panel-level processing My goal is to put the right names on my radar in hopes of finding the next $MP, $AXTI, $OKLO. Let me know below what other baskets and/or stocks I should put on my radar.

@ChartsJavi @grok have they had a successful launch? Didn’t their last one in Bertram explode?

$FLY keeps trending

@fundmyfund This was a damn nice read easy on the eyes.

$AMPX Q3 ER - this was a great report. $21.4M v $7.3M a year ago. +173% YoY; 42% sequential. Analysts were at $16.7M not that I care too much about their opinions but the market does. This is what we want to see in our young high growth companies. Reward? The stock is indicating a paltry +4% in pre as I type this. In early October this would have been +30% easily. Environment matters. I did not have time to listen to the conf call last night; that will be on the docket shortly. This company is rocketing up the food chain in my focus area as it executes. It's also a backdoor play on drones. 75% of sales were UAS. It also is a grift play - the taxpayer is paying for part of its capital needs. The only bugaboo is valuation but it is far cheaper than say $ENVX and high growth companies can be highly valued. Ask Alex Karp. (ask him gently, he tends to get agitated) We removed a $7M quarter a year ago and added a $21M quarter. So trailing price to sales plummeted quite a bit. What is the price for a 100% grower? It should be high. p.s. $LPTH will be growing like this "sooner than you think". With a far cheaper valuation. But I digress! I love their earning reports format - so many companies could learn from $AMPX. Not only do they go in depth on the quarter they throw some education along the way. 8 of the 14 pages are prose, before they get to data and financial tables. Bravo. 👏 Will update these notes when I listen to the conf call. 🎧📞 ********************* So many positive metrics outside the outstanding revenue growth. Let's count: 1) Q3 159 customers received shipments. 80 were new customers. Of 444 in their rolodex. "n the third quarter, we shipped batteries to 159 end customers, 80 of whom are new to the Amprius platform. The remaining 79 are repeat customers. We don’t ship to every customer in every quarter, but we do expect to gain new customers every quarter" "Looking at our customer base, about 75% of our revenue in the third quarter came from the aviation segment, led by the unmanned aerial systems (UAS) market. The remainder of our Q3 revenue was primarily derived from the light electric vehicle (LEV) sector, which remains healthy but has a lumpier profile due to our customers’ varying product introduction cycles. The LEV market tends to have shorter design-in cycles, and we believe our drop-in replacement batteries can help us succeed in gaining market share in this growing market." 2) Mega order for a co this size "One of our major wins this quarter was a record $35 million purchase order from a leading UAS manufacturer, which we announced in September 2025. This order is a follow-on purchase from the same customer that placed a $15 million order earlier this year." 3) $AVAV relationship, who is also rocketing up my food chain as they diversify their product line. "During the third quarter, we also deepened our relationship with another key customer, AeroVironment. As part of the U.S. Army’s xTech Prime program, we shipped samples of our ultrahigh-energy cells for evaluation in a variety of applications. These cells reach up to 520 Wh/kg and vastly improve endurance, payload capacity, and mission economics for high-altitude platforms." 4) They work with Vikings. "Another key Amprius partner in the drone segment is Nordic Wing in Denmark. In Q3, they chose our SiCore cells to power their Unmanned Aerial Vehicle (UAV) platform after an extensive qualification and evaluation period. The Astero ISR with Amprius SiCore batteries flies 90% longer than with standard cells. Astero stays airborne longer, covers more ground, and delivers real-time intelligence with fewer interruptions." 5) Maybe doing something with $AMZN down the road "Looking a bit further out, we continue to make significant inroads in our relationship with Amazon. After being selected for the inaugural Amazon Devices Climate Tech Accelerator cohort in July 2025, we successfully advanced to the integration assessment phase." 6) Grift engaged "As we previously discussed, we are working closely with the Defense Innovation Unit (DIU). Our DIU contract gives us funds to increase the capacity of our Fremont, California pilot line to 10 MWh and expand our capabilities to support quick turn SiCore customer prototypes. Since our last update, we have received an additional $1.5 million follow-on contract, bringing our DIU contract total to $12.0 million. Our program mandate includes qualifying individual lithiumion battery components from National Defense Authorization Act (NDAA) compliant suppliers, which will allow us to work more seamlessly with the Department of Defense. This effort is part of a large momentum shift to US domestic production of batteries." 7) They use contract mfg (some people don't like this, I like asset light businesses) "As we progress on building out an NDAA compliant supply chain and production capacity for customers that require it, we have continued to utilize our contract manufacturers to support our rapid growth. As a reminder, we have over 1.8 GWh of capacity available to us through our partners, including our most recent added partner in South Korea. This capital-efficient model provides productiongrade cells for qualification today and supports our ramp to volume while still allowing configuration control and aerospace-aligned quality systems. We are also opportunistically sourcing additional partners to provide us with greater geographic diversification and operating flexibility." 8) Already momentum in Q4 "We are carrying this momentum into the fourth quarter. A few weeks ago, we announced that ESAero, another leading UAS company, chose our SiCore SA08 cells to power their group 1 and group 2 UAVs that support defense, security, logistics, and public safety applications" 9) Lack of customer concentration "Within our customer base, only one customer accounted for more than 10% of our revenues in Q3. Going forward, we plan to continue adding to our customer mix to diversify our revenue base and we believe that as we develop more diversified contract manufacturing capacity, additional demand may be unlocked." 10) Backlog growing! To be fulfilled "near term" so I'd expect that in next 1-3 quarters. "At the end of Q3, we had $53.3 million of orders, including the $35 million order from a UAS manufacturer, to be fulfilled in the near term. This backlog is 83% higher than the second quarter and it highlights our team’s ability to drive demand." 11) Gross margin % low but growing as they scale. Last Q was their first positive GM I believe (will check notes) - will listen to conf call if they give a target gross margin down the road, hopefully 35%+. Don't know that info yet. "Our cost of goods sold was $18.1 million in Q3 and it did not increase at the same rate as our revenue due to a favorable product mix and higher volumes. This in turn increased our gross margin to 15%, a significant improvement over our gross margin of 9% in the second quarter." 12) No debt "We exited the third quarter with $73.2 million in cash and no debt." 13) Cash burn is about $10M per Q so no immediate need to dilute . If they don't go cash flow positive by 2027 they will need to dilute. "$9.2 million used in operating cash flow, which was flat with our average projected run rate of about $2.5 to $3.0 million monthly, excluding transaction-related costs. The primarily drivers this quarter are the improvement in our net loss offset by the increase in accounts receivable at the end of the period due to our increase in sales. • $0.4 million used in investing activities related to our Fremont, California facility" 14) Taxpayer pays for our capital investment - grift baby grift "We have made the decision to strategically invest in diversifying our supply chain and expanding manufacturing capacity within our Fremont facility to include electrode manufacturing. We are doing this in collaboration with the U.S. Government Defense Innovation Unit and have secured a contract for $12.0 million awarded in the third quarter. With what we know today, we expect this funding to cover the majority of our capital investment in the next several quarters as we work to develop a growing and resilient source of supply in a dynamic trade environment." ************************ My financial model utilizing a 12 price to sales on 2026 estimated revenue says $AMPX is fairly valued here. That said, I don't know if $105M is conservative for 26 sales (it feels so). They just put out a $20M+ quarter and have backlog that is >50% of my 2026 number. Not sure how much of the backlog hits Q4 25 vs 2026. I need to do more homework. Also if they can grow in excess of 40% the next few years, 12 price to sales is conservative. A lot of moving parts. But an exciting opportunity. d1io3yog0oux5.cloudfront.net/_e34561065916b…

@DudespostingWs @grok who makes these and in US as well?

@aleabitoreddit You already commented on $ALMU. What about $AMPX, $AEHR, $ACMR ?

January is wild. Not even joking, many of my picks are up 60%+ over the past few days. Didn’t even include $AXTI +60% or $VLN +60% in the charts. While leaving out smaller rallies like $CRDO 25%+ or Intel 25%+. Need more 5x stocks for research if you have ideas or a thesis.

@boring_invest $ALMU Interview youtu.be/H8iIkQGD1Ds?si…

I wanna focus on the CFO of $ALMU for a minute. He has an impressive track record. As per grok: Stewart brings more than 20 years of experience in financial leadership, primarily at public and high-growth technology companies. His career highlights include: - Serving as CFO at LeddarTech Holdings Inc., an AI-driven software company, where he helped guide the firm through its public market debut. - Prior to that, as CFO at Bionano Genomics, Inc. (BNGO), during which the company achieved a 62% compound annual revenue growth. He joined Bionano in September 2020 and contributed to its expansion in the genomics field, providing tools for researchers and diagnostics. - Earlier roles included CFO at Microchip Biotech, a venture-backed single-cell analysis instrumentation developer that was acquired by Standard BioTools Inc. (LAB). - Senior financial positions at Maxwell Technologies (acquired by Tesla), Entropic Communications (acquired by MaxLinear), and Intel, showcasing his involvement in multiple successful acquisitions and growth phases. A CFO like that would definitely weigh the odds of success before joining a company. I trust his intuition on $ALMU and believe he wouldn’t join if he didn’t view the odds of a successful commercialisation as very favourable.

@BlackScholesMan Thanks I had to jump off. You covered Korea and euro and Japan& ships, my only one I didn’t get to hear about was BAE which has a lot of business here in the US Recent follower and I wrote down 10+ name of yours, but unfortunately, they’re all high as a cats back at this pt

@robptrsn At some point yes

@MartyChargin EWY to get Hynix and Samsung

$EWW $EWY $FEZ $ECH ... These foreign markets continue to deliver the goods....

@AKWilk I would argue. It might be one of the best American and global geopolitical history books you’ll ever read.

Just started. Solid so far.

@Pref_Right @DanielYergin Congrats. It is a must read. Took me 10 years as well. One of best history books anyone could read. Should be a requirement for energy folks. The film version is crap btw. If I may suggest another history book, the Ascent of Money by Niall Ferguson

Finally finished @DanielYergin “The Prize”. Admittedly, I began reading December 2012. Started over in September ‘23 and I really enjoyed it. It’s like 750 pages, so not easy to stay focused. But it’s the best, most comprehensive, book of industry I’ve ever read. Loved it.

@Pref_Right You don’t know me but you know my son Walt. He says hello

@robptrsn Used to be a naturally aspirated v8 plus solid axels to cut down on things that could go wrong. Bulletproof out of the factory.

I feel like the new Land Cruiser didn’t need to be so extreme in its desire to compete with Jeep and the Bronco. Those design decisions show me in either totally out of tune to the suv market, I’m a niche customer, or Toyota is out of place. And Toyota rarely misses.