JasonEllis

i said it before and i’ll say it again, hynix is a 2.5-3T company..

Trio-Tech ($TRT): My Next +100% Bag this Year. My investing philosophy for 2026 is simple: find the companies that own the un-bypassable infrastructure, buy them cheap, and ignore the daily noise. Trio-Tech ($TRT) fits that description perfectly. They don't sell hyped-up promises; they run physical back-end labs in Southeast Asia that the tech elite depend on. After their latest 10-Q filing on Thursday, the stock jumped 40% in a day, locking in a +100% gain for my port. Let’s look past the sudden price spike and break down the actual balance sheet. Here is why Trio-Tech International ($TRT) just stopped being an ignored micro-cap and turned into a critical, un-bypassable infrastructure lynchpin for the second half of 2026. ➡️This stock has delivered another +100% for me this year in a relatively short period. I initiated my first position 23rd April and accumulated on the way up, with a cost basis of ~$10. I’m not selling a single share. 👇 The Q3 2026 Earnings Anatomy: Numbers Without the Makeup On May 14, 2026, Trio-Tech published its Q3 FY2026 Form 10-Q on the NYSE American exchange. The market's reaction was violent - the stock surged over 40% in a single session, tearing past the $20 mark. To understand if this rally has actual fundamental legs, we need to strip away the market euphoria and dissect the raw financial architecture. 1⃣Top-Line Explosion & Segment Shift TRT generated $16.51M in revenue for the quarter (ended March 31, 2026) compared to $7.38M in the prior-year period. That is a staggering 124% YoY growth. The true magic, however, lies in where this money is coming from: ➡️Semiconductor Back-End Solutions (SBS): $13.08M vs. $5.43M last year (+141% YoY) ➡️Industrial Electronics (IE): $3.43M vs. $2.00M last year (+76% YoY) Nearly 80% of TRT’s entire business is now concentrated in the SBS segment - specifically final reliability testing and "burn-in" services for packaged silicon. This is a direct consequence of the insatiable demand for physical validation of AI data center clusters and 800V automotive grids. 2⃣The Gross Margin Paradox: Risk or Scaling Feature? The bears and basic trading algorithms immediately flagged one metric: gross margins compressed sharply from 27% down to 16%. In the micro-cap world, a sudden margin drop often triggers panic. But a granular look at the balance sheet reveals this is a temporary byproduct of massive scaling, not structural decay. The compression is driven by two highly logical factors: ➡️A deliberate transition into massive, high-volume testing contracts which carry lower unit margins but secure immense aggregate dollar volume. ➡️Mobilization & Ramp-up Costs: TRT is aggressively preparing its operational footprint for an unprecedented influx of business. The engineering, logistics, and setup costs are being recognized today, before the new facilities book their first dollar of revenue. ⬇️ Despite this margin pressure, the net loss for the quarter was narrowed to just $38k (compared to a loss of $495k in Q3 FY2025). The company essentially achieved an operational break-even while expanding at a breakneck pace. 3⃣The Malaysian Masterstroke: Embedding into the Penang Megacluster The absolute crown jewel of this investment thesis is buried in the company’s regulatory filings from May 2026. Trio-Tech signed a lease agreement for a massive 104,000 square foot facility in Perai (Penang, Malaysia) at a cost of $115k per month, effectively commencing June 1, 2026. Let's put that footprint into perspective: ➡️Doubling Global Capacity: This single facility increases TRT’s global laboratory testing square footage by an estimated 80% to 100%. ➡️The "China+1" Geopolitical Arbitrage: Western chip designers are desperately trying to de-risk their supply chains by moving back-end processes out of China. Penang has established itself as the undisputed capital of semiconductor packaging (OSAT) in Southeast Asia. By placing a 104k sq ft lab right in the heart of this industrial zone, TRT is setting up shop door-to-door with a legendary roster of tech titans. In fact, $AMD, $INTC, $AVGO, $JBL, $COHR, and $KEYS are all located within a tight 15-to-20 minute drive from new $TRT location. Instead of packing highly sensitive, multi-thousand-dollar AI modules onto planes to be tested in mainland China or Taiwan, Western tech companies can now complete the entire packaging and final validation loop locally in Penang. TRT is building the toll booth right outside their factories. 4⃣The Proprietary Moat: Vertically Integrated Hardware Most testing service companies face a brutal operational bottleneck: they have to buy their burn-in ovens and test equipment from external vendors, forcing them to pay immense markups and wait in line for tool allocations. Trio-Tech operates on a fundamentally superior architecture. They design and manufacture their own burn-in systems, custom Burn-In Boards (BIBs), and hydraulic centrifuges in-house. This creates a dual advantage: They don't wait for supply chains. They can outfit the massive new Penang facility at internal manufacturing cost, dramatically lowering their CapEx requirements. They do not sell this equipment for a one-time product margin. Instead, they deploy it internally to sell recurring, high-volume testing services to the biggest names in tech. This is where the massive operating leverage comes into play. The lease and fixed overhead costs for the Penang facility are static. Once the utilization rate of their in-house manufactured machinery crosses the 75% threshold, every additional hour of chip-cooking translates into pure, exponential profit falling straight to the bottom line (EPS). 5⃣A Fortress Balance Sheet Meets Undervalued Multiples In April 2026, the company executed a registered direct offering, securing $10M in gross proceeds from institutional investors at $9.50 per share. Retail investors initially lamented the dilution. In reality, this capital was immediately deployed as the working capital engine required to stand up the Penang mega-lab and execute on their surging backlog, which includes a confirmed $7.8M in firm orders for specialized BIBs tailored exclusively for next-generation AI GPU architectures. When you look at the balance sheet post-offering, the financial health is pristine: ~$28.3M in total net cash (combining March cash balances with the April raise). Long-term bank debt: A negligible $255k. ⬇️The Valuation Disconnect Following the 40% post-earnings surge to around $21, and accounting for the expanded share base (now sitting at roughly 5.1M shares outstanding), TRT's market capitalization rests at approximately $202M. Over the first 9 months of fiscal 2026, TRT has already locked in $47.7M in revenue. Even under a highly conservative assumption that the fourth quarter simply matches the current run-rate ($16.5M), the company will comfortably close the fiscal year with $64M in revenue. This leaves the company trading at a forward EV/Sales multiple of 3x. Paying a 3x Enterprise Value-to-Sales multiple for a debt-free, asset-owning business that is actively doubling its global physical capacity, while anchoring itself right in the middle of a 15-minute radius shared by AMD, Intel, and Broadcom - is a textbook market mispricing. At a $10 cost basis, this position offers an immense margin of safety. The structural fundamentals are moving significantly faster than the stock price. I am locked in for the long haul, and any macro-driven volatility or short-term pullbacks will be viewed as an open invitation to accumulate more shares.

Marlon shares an old clip from before he blew up, where he promised himself he would stay consistent, get his visa, move to the States, and get a house. 2 years later, he really made it happen 🥹

方向是对的，AI光化是物理必然。 但几个细节需要纠偏： NVL72机架光占比远低于20%。GB200最大的工程亮点恰恰是机架内Scale-Up用无源铜缆，省下20kW功耗。这是Jensen亲自在GTC讲过的事。光的战场在跨机架，不在机架内。 写光通讯不提 $AVGO 和 $MRVL，等于写存储不提 $MU。这两家才是CPO时代的真正受益者。AVGO在交换机芯片+CPO研发第一梯队，MRVL在光模块DSP接近垄断。 硅光和传统InP长期是替代关系。把 $LITE 和 $TSEM 同时重仓，意味着你既赌InP继续主导，又赌CPO快速量产。这两件事不能同时为真。 我的看法不变 光通讯有真实的alpha。但alpha集中在已经被市场认可的leader上，不在被精心挑选的”低估值7龙珠”里。 我现在盯的方向： $LITE $COHR 是Tier 2 其他暂时观察 不追故事，等setup

Memory and power shortages will worsen before they get better If you sit down and analyze the tech trends, you quickly realize there is no way supply can catch up with demand Only massive demand destruction could bring prices down And the AI capex cycle is backed by the deepest pockets on Earth hypertechinvest.com/p/portfolio-up…

My friend GSR wrote a really good piece. He found a new bottleneck. The market is overly focused on indium prices, but the more structural bottleneck in the InP supply chain is qualified high-purity phosphorus. NCI and Rasa are key Japanese suppliers of red phosphorus for InP feedstock, or 6N/7N high-purity red phosphorus. Once again, obscure Japanese companies I had never heard of are controlling the bottleneck lol. This material is not just a simple commodity input. It is a process-sensitive material directly tied to crystal growth yield and qualification.

Rubi Rose had Nina Drama BLUSHING because she genuinely couldn’t believe how YOUNG she looked despite being 37 years old 😭❤️ “How old are you?… You’re lying b*tch… God bless you. You look fantastic… You’re maybe 24, 25.”

I talked to one of the better optics / photonics analysts today... his top 2 picks are $LITE (1st) and $COHR (2nd)... he said $AAOI is one of the riskiest in his group (because of management credibility) but says $AAOI definitely has the most upside from current prices if management is able to hit their aggressive financial targets. NFA DYOR **We own $LITE $COHR and $AAOI at @FirstWaveFund

⚡️ President TRUMP just BOUGHT these AI stocks that I am personally BULLISH on: - Photonics: $JBL, $COHR, $LITE - SaaS: $NOW, $PLTR - Memory: $MU, $PENG - Infrastructure: $DELL, $ORCL, $CRWV, $VRT - Chips: $AMD, $INTC, $NVDA - Energy: $BE, $GEV Top 3 would be JABIL INC., DELL TECHNOLOGIES & PENGUIN SOLUTIONS 🔥

Ex Google CEO, Dr. Eric Schmidt: AI may hit a money wall before it hits a power wall. "The real limit to AI is not energy; it is actually cash. When you add up the cost of these things, if you take round numbers, say $50 billion per gigawatt, then 10 gigawatts is half a trillion dollars. How many companies, countries, and so forth can hand an industry a trillion dollars of capital? Very, very few. The Chinese could certainly do it. I do not know if they are doing it, but I am going to try to find out. In America, there are people who hope that is going to happen. It is interesting that you can finance these things because the brilliance of the American capital market allows us to borrow that kind of money. For example, the Europeans cannot do this, which they are sort of sore about." --- Full video from 'Special Competitive Studies Project' YT channel ( link in comment)

the obvious bear case on photonics is: capacity is coming. but....does it comes fast enough? no. the demand-supply gap will last longer than most believe 3 reasons: - CPO demand is ramping faster than InP fabs. Goldman sees optical networking going from ~$15B in 2026 to ~$154B by 2028, with CPO contributing ~$91B. Every CPO architecture still needs external InP laser sources. - The bottleneck is moving upstream. LITE, COHR, AAOI can add device capacity, but they still draw InP substrate wafers from a narrow supplier base: AXT/Tongmei, Sumitomo, JX Advanced Metals. Today’s laser constraint can become 2027’s substrate constraint. - Fabs do not scale by spreadsheet. Reactors, cleanrooms, process conversion, yield ramp, and qualification take 18 to 24 months minimum. Supply is coming. But AI networking demand is coming faster. That is why photonics is still trading like a bottleneck. chart source: Goldman's optical networking report

@FrontRowBrian @MDougy7 @arielhelwani Then say it wtf?? Change your wording.

@MDougy7 @arielhelwani I would have said 'killed'

This guy murdered a young woman like a year ago and no one asks him about it. And it happened in the most corrupt shit hole country on earth. But no one will ask. Where’s the great journalist @arielhelwani ?

@retail_mourinho What's wrong with everybody? $1k to $15k is a penny in his portfolio lmao

Did Trump really just buy $ASTS?

The White House is telling you exactly where to invest…don’t ignore them $DELL $AVGO $SNPS $NOW $ADBE $CRM $NFLX $MSFT

This was so good. I read the whole thing. @insane_analyst has so much knowledge

Took a small position in $PENG this week $PENG: $2.5B mcap AI factory platform. SK Telecom $200M anchor + SK hynix trilateral JV locks in HBM roadmap. CXL 2.0 NV-CMM live, Optical Memory Appliance late 2026 attacks HBM scaling chokepoint via optical I/O. FY26 guide raised 6% → 12%. 85k+ GPUs deployed, sticky Scyld/ICE orchestration software. Just feels cheap compared to the rest of the market here? $PENG

Three ETFs Targeting the Next AI Infrastructure Bottlenecks The first wave: own the obvious AI leaders. But investors may now need to ask where the bottlenecks are forming. Three ETFs offer a useful framework for this shift: SMH, DRAM, and EUV. Each targets a different layer of the AI infrastructure chain. SMH gives broad semiconductor exposure, DRAM isolates the memory bottleneck, and EUV targets lithography, photonics, and optical infrastructure. SMH is the most institutionalized option. It provides exposure to large semiconductor and semiconductor-equipment companies rather than one narrow choke point. Key holdings include $NVDA at 17%, $TSM at 10%, $AVGO at 8%, $INTC at 8%, $AMD at 7%, $MU at 6%, $TXN at 5%, and $KLAC at 4%. This makes SMH the most natural core holding of the group for investors who want exposure to the full AI hardware stack. It covers GPUs, foundries, custom silicon, CPUs, memory, analog chips, and semiconductor manufacturing tools. The trade-off is lower purity. SMH is not a single bottleneck bet. It is a broad semiconductor ecosystem bet. DRAM is more targeted. It is designed around the AI memory squeeze, with exposure to HBM, DRAM, NAND, and storage demand. The fund is highly concentrated. Its largest positions are SK Hynix at 28.15%, $MU at 27.16%, and Samsung at 19.67%. Together, these companies dominate the global memory supply chain. Smaller holdings such as Kioxia, $SNDK, $STX, $WDC, Nanya, and Winbond add exposure across NAND, SSDs, HDDs, and specialty memory. DRAM is arguably the cleanest expression of the AI memory bottleneck. It is also more momentum-driven and concentrated than SMH, with a higher 0.65% expense ratio. EUV is the most specialized and higher-risk ETF in the group. It focuses on the “light layer” of AI infrastructure: photonics, EUV lithography, optical networking, semiconductor inspection, and precision manufacturing tools. Holdings include $TSM at 9.52%, $ASML at 7.97%, $GLW at 5.19%, $LRCX at 4.98%, $AMAT at 4.84%, $LITE at 4.46%, $CIEN at 4.32%, and $KLAC at 4.07%. AI data centers increasingly face limits around power, bandwidth, packaging, and interconnect speed. Photonics and advanced lithography may become critical as compute demand scales. Framework: SMH = core AI semiconductor exposure DRAM = memory bandwidth and capacity bottleneck EUV = lithography, photonics, and optical infrastructure bottleneck

The director for this movie had some 10/10 Jill valentine casting in his hands but couldn't resist making 90% of the RE movies about his OC played by his wife