HBrsn

The photonics thesis is predicated on a multi-year cycle. My play? Focus on two concentrated trades that account for both short and long term technology shifts so I can gobble up the asymmetry today and tommorow. The first, $AAOI, is my near-term explosive set up based on extreme hyperscaler demand for LPO transceivers to solve the copper wall now. Management is guiding to a $5B run rate as soon as early 2027, up from ~$500M in 2025. MC ~8B pales compared to the outlook. Vertical integration with their own domestic laser fabs makes it unique. I know the demand is there. This is not a sleep at the wheel trade. It means watching management like a hawk to see if they can meet the moment. The second, $SOI / $SLOIF, is a trade on the back half of the cycle, when CPO takes over because transceivers are imperfect solution. A pluggable transceiver or LPO uses approximately 50 square millimeters of Photonics-SOI per chip. A CPO switch package uses approximately 200 square millimeters. This business is cheap at 3.6x EV/revenue because it will take time to bounce off the mobile chip trough and let photonics point the way once CPO production ramps in 2027. Morgan Stanley sees this doubling their PT yesterday to 70 euros, considerable upside on 54 euros currently. "The mobile market remains weak, but growing confidence in Photonics SOI is shifting the medium-term outlook". Soiboy and AA-O-my-god.

@jasons_chips #soitec

@jasons_chips #soitec

I publish my first article on Soitec $SOI $SLOIF 2 weeks ago. It was very popular. The market has begun realizations and the stock went up 25%. Now the widely anticipated part 2 is here. And again it is free. Soitec will soon have a complete and utter monopoly on silicon sandwich wafers. Learn about why below. We cover everything from the 7 step process for making SOI to specific market share numbers and cheeky management comments. If you’re invested in AI/photonics this is must read.

$SOI SOITEC Likely Current Customers based off geography mix Singapore (~50% of H1 revenue — €115.5m): This is almost certainly GlobalFoundries. GF operates one of the world's largest fab complexes in Singapore (Fab 7) and is the dominant RF-SOI foundry globally. The sheer concentration — half the company's revenue from one country — strongly suggests GF is the single largest customer, probably by a wide margin. GF is the go-to foundry for Qualcomm, Skyworks, Qorvo, and others on RF-SOI process nodes. UMC also has Singapore fabs and does some SOI work, but likely a much smaller piece. France (~31% — €71.4m): STMicroelectronics — Crolles fab is literally down the road from Soitec in Grenoble. ST is the highest-profile adopter of FD-SOI globally and has built multiple process nodes around it. This is almost certainly the bulk of France revenue. Possibly some X-FAB or internal qualification volume, but ST dominates. United States (~12% — €26.7m): Two likely customers here. Tower Semiconductor (Newport Beach fab) — your confirmed biggest SiPho foundry customer, so this is the photonics-SOI revenue showing up. And Skyworks directly — the doc references a multi-year POI supply agreement for the SKY5 platform. Possibly some volume to Intel Foundry Services if they're pulling any SOI. Other (~7% — €17.2m): Likely a mix of TSMC (Taiwan — COUPE platform, probably still early/small), Japan (Shin-Etsu royalties or direct sales to Japanese IDMs), and possibly some Samsung or Dongbu HiTek volume.

I can say with high confidence I was the first to spot $SOI Soitec’s monopoly and call it out on X and Substack. When I started writing about it, the response was minimal, I never imagined that just four months later there’d be “Soiboys” on X.

포토닉스 섹터 폭락의 의미와 매수 기회ㅣ260327 1. 오늘 포토닉스 섹터 전체가 동반 급락했음. $LITE (루멘텀), $COHR (코히런트), $SIVE (시버트), $MTSI , $AAOI 같은 레이저 기업들, $IQE , $AXTI (AXT), $SOI , $TSEM 같은 기판·파운드리·에피웨이퍼 기업들까지 줄줄이 빨간불이 켜졌음. 낙폭도 베타가 낮은 $COHR 6%대부터 일부 종목은 22%까지로, 섹터 전반이 폭격을 맞은 수준임. 이걸 이해하려면 왜 이렇게 강한 섹터가 하루 만에 이렇게 빠지는지 구조부터 봐야 함. 2. 오늘 하락은 실적 악화나 사업 펀더멘털 붕괴가 아니라 매크로 유동성 충격임. 매크로 리스크(금리 변동, 트럼프 관세 불확실성, 지정학 긴장 등)가 터지면 투자자들이 레버리지 포지션을 강제로 청산하거나 스탑로스가 연쇄 발동되면서 "유동성 진공(liquidity vacuum)" 상태가 생김. 이때 평소에 잘 나가던 고베타 종목들이 가장 먼저, 가장 깊이 빠짐. 오늘 포토닉스 섹터가 정확히 그 패턴을 밟은 것임. 3. $LITE (루멘텀)은 올해 초부터 극적인 스토리를 써왔음. 3월 초 NVIDIA가 $LITE 와 20억 달러 규모 멀티이어 광학 부품 공급 계약을 체결하고 전환우선주까지 인수했으며, S&P 500 편입(3월 23일)까지 확정됐음. 실적도 EPS 1.67달러(예상 1.41달러)로 어닝비트, 매출 6억 6550만 달러에 전년 대비 65.5% 성장을 기록했음. 그럼에도 내부자(CFO)가 3월 초 380만 달러어치를 매도했고, P/E가 238배까지 뻥튀기돼 있던 밸류에이션이 오늘 하락의 빌미를 줬음. 4. $COHR (코히런트)는 포토닉스 섹터에서 낙폭이 상대적으로 낮은 6%대에 그쳤는데, 이건 종목의 펀더멘털 탄탄함을 방증함. $COHR 는 OFC 2026에서 타워반도체와 협력해 400Gbps 레인당 실리콘 포토닉스 기술 돌파구를 선보였고, 3.2T 트랜시버 아키텍처의 핵심 공급자로 포지셔닝됐음. AI 데이터센터 광학 시장의 2030년 SAM(서비스 가능 시장)이 200억 달러로 추정되는 상황에서 $COHR 는 InP 레이저, CPO 솔루션, 200G EML 등 전 스펙트럼을 수직 통합 공급하는 거의 유일한 플레이어임 5. $SIVE 는 오늘 가장 주목해야 할 이름임. 현재 시가총액이 약 3억 4000만 달러로 극히 낮은 소형주이면서, CW(연속파) 레이저 분야의 다음 주자로 언급되고 있음. $COHR 가 InP 레이저를 2027년까지 완판 상태로 공급이 막혀있는 상황에서, $SIVE 가 그 수요를 받아낼 다음 병목 공급자로 부상할 수 있다는 논리임. 시총 3억 4000만 달러에서 수요 전환이 확인되는 순간 멀티배거 가능성이 열리는 구조임. 6. $AXTI (AXT)는 기판 섹터의 핵심 병목 기업임. $AXTI 는 전 세계 인듐포스파이드(InP) 기판 시장에서 약 40% 점유율을 보유하고 있음. 올해 초 중국 수출 허가 지연으로 Q4 2025 매출이 22% 급락했지만, 2026년 상반기부터 수출 허가가 정상화되기 시작했고, 하반기에는 InP 생산 능력을 두 배로 확대할 계획임. 연초 대비 YTD 수익률이 무려 223%에 달했던 종목이니, 오늘 낙폭은 수익 실현과 매크로 쇼크의 합작임 7. 오늘 폭락이 "손 바꾸기(exchange of hands)"의 기회라는 표현이 핵심을 찌름. $LITE 와 $COHR 의 InP 레이저는 2028년까지 완판 상태라는 사실은 오늘 주가가 빠졌다고 해서 바뀌지 않음. 수요가 사라진 게 아니라 레버리지 투자자들이 강제 청산당하면서 주식이 싸게 매물로 나온 것임. 이런 날이 오히려 장기 투자자가 가장 유리하게 포지션을 재구성할 수 있는 날임. 8. 레버리지를 쓰지 말라는 교훈이 오늘처럼 잘 맞아떨어지는 날도 없음. 하루 25% 오를 수 있는 종목은 하루 20% 내릴 수도 있음. 레버리지 없이 현물로 보유한 투자자는 오늘 하락을 그냥 견디면 되지만, 레버리지를 쓴 투자자는 스탑로스 발동으로 최저점에서 강제 청산당하고 이후 반등을 구경만 하는 최악의 시나리오를 겪게 됨. 변동성 자체를 적으로 볼 게 아니라 구조적 우위를 가진 플레이어에게 주어지는 할인 쿠폰으로 봐야 함. 9. 지금 이 섹터에 어떻게 접근할지 구체적으로 정리하면 이렇게 됨. 첫째, $COHR 처럼 낙폭이 상대적으로 작고 펀더멘털이 검증된 종목은 지금부터 분할 매수 구간으로 설정할 수 있음. 둘째, $SIVE 처럼 소형주이면서 다음 수혜 포지션이 예상되는 종목은 전체 포트폴리오의 극히 일부로만 소량 진입하고 변동성을 감수해야 함. 셋째, $AXTI 는 하반기 생산 능력 확대 일정이 구체적이므로 2026년 하반기를 겨냥한 중기 포지션으로 분류할 수 있음. (이어서 계속 🔗)

This crash today is a good reason why it's good to stay invested. -> $AAOI is still up 233%? -> $IQE is still up ~100%? -> $SIVE is still up 100%+? -> $LITE is still up ~88%? -> $TSEM is up ~48%? -> $SOI is still up 36%? Even after today my $AXTI positions are still up ~680% after the drop today? Yes, red days look rough if something drops 13% or 20% and especially so if you're later to enter. But if you keep thinking it's going to: -> drop from $500m and sit out -> drop from $1b and sit out -> drop from $2b and sit out -> drop from $3b and sit out By the time it's $4B and there's finally a correction 15%, you missed out on the entire rise. People can say "hindsight bias, you're posting after it went up". No, I've posted these are the returns after my original thesis, and I'm just riding the wave up. I keep getting trolls like these in my comments sections, but hope it doesn't make others panic. Not everything moves in a green line up, same applies to other sectors and stocks if there’s no material change. Risk exists. And portfolio sizing is very important. But if you’re going to enter higher beta names in bottlenecks especially with a clear thesis in mind. Need to learn to embrace the volatility for your thesis to play out.

@PhotonBull #soitec

Why is the timeline obsessed with laser suppliers like $LITE when the most impenetrable AI monopoly is the atomic scalpel that builds their starting wafers? Routing light for AI networks requires exactly 220 nanometers of silicon sitting perfectly on a 2-micron oxide layer. If the silicon thickness varies by even a single nanometer, light leaks into the base material and the optical signal dies. Conventional polishing methods physically cannot achieve this sub-nanometer uniformity at volume. The only industrially viable answer is Soitec's Smart Cut. Smart Cut is a proprietary layer-transfer process that implants hydrogen ions into a wafer to form microscopic bubbles. When heated, these bubbles cleanly cleave flawless, atomic-level layers of silicon. As AI networks migrate to Co-Packaged Optics to survive a severe thermodynamic power crunch, this specific substrate technology undergoes a massive 4x wafer content multiplier, jumping from 50mm² to 200mm² per port. Here is the sole company positioned at this structural chokepoint: Soitec $SOI: This hidden monopoly owns the Smart Cut process with over 4,300 active patents. They control over 95% of the photonics substrate market, protected by a brutal 12 to 24-month foundry qualification lock-in that practically eliminates all competition. $SOI $SLOIF

@MitchMartan98 100% $SOI #soitec

After all this carnage is done with, there will be an inevitable rally at which point many will say... "I wish I bought more of X down here" What stock are you looking to add? I have -> $RDDT $APP $BE $SOI / $SLOIF $HROW and $NBIS on my radar.

This deep-dive report is the perfect introduction to the market's most misunderstood hidden monopoly: Soitec. Soitec manufactures the engineered Photonics-SOI substrates that serve as the molecular foundation for nearly every silicon photonics chip on earth. Using their highly proprietary "Smart Cut" technology, they command an estimated >95% market share in this indispensable niche. While the stock has been beaten down by a cyclical slump in its legacy mobile communications segment, its AI photonics division is quietly compounding at a 25% annual growth rate. The industry-wide push toward Co-Packaged Optics (CPO) is the ultimate catalyst that will explode their top line. A traditional pluggable transceiver requires about 50 square millimeters of Photonics-SOI. A CPO switch package demands roughly 200 square millimeters. That is a massive 4x hardware content multiplier for every single optical port that upgrades to bypass the physical limitations of copper. Institutions are finally waking up to this exact math and the resulting supply crunch. Morgan Stanley just doubled their price target on the stock today, aggressively bumping it from €35 to €70. Link in bio. $SOI

@athuinvests #soitec

Photonic stocks just got slammed today. Here are 3 stocks in my portfolio and why they stand out for me: - Soitec ( $SOI / $SLOIY ): They own the silicon-on-insulator (SOI) wafer market - the foundational substrate for next-gen silicon photonics. Western SiPh monopoly at ~2.4x P/S. When CPO moves from pilot to production inside AI racks, these wafers go from niche to must-have. Pricing power is real here. If the inflection hits as expected, this valuation looks absurdly cheap in hindsight. - Sivers Semiconductors ( $SIVE ): Pure-play laser supplier for the CPO wave. They make the critical CW DFB lasers that light up high-speed optical I/O. Direct exposure via partnerships with Ayar Labs (Amazon Trainium pipeline) and POET (Marvell/Microsoft Maia ties), plus Jabil already shipping their tech in 1.6T transceivers to Amazon, Meta, etc. Market cap just dipped to ~$330-340M after a 22% drop. Still early, still underfollowed, massive asymmetry if CPO adoption accelerates. This is the kind of name that can rerate violently once the orders flow. - Applied Optoelectronics ( $AAOI ): Already past the “story” stage - they have real revenue and manufacturing scale today. The bull case is a multi-year optical component ramp tied directly to AI infrastructure buildout. If their 2027 revenue trajectory even partially materializes (8-10x growth potential), today’s valuation will look comically compressed. AI scaling is no longer bottlenecked by energy, compute or memory alone. The interconnect layer is the new chokepoint. Photonics solves it. The hyperscalers are already voting with their wallets. The market hasn’t fully repriced this foundational shift yet. Today’s blood-red tape in photonics? It’s not a warning - it’s a handover. Volatility is the feature, not the bug. The names with real technology, real customers, and real scale will be the ones left standing when the next leg of the AI buildout kicks in. Position accordingly. NFA. DYOR. But if you’ve been waiting for a clean entry into the CPO/SiPh supply chain… the dip just served it up.

@PhotonBull #soitec

The traditional data center is dead; the era of the AI Factory has arrived. AI factories are a new class of extreme-scale, GPU-driven facilities designed specifically for generative AI and large language model workloads. Instead of isolated servers, these facilities utilize massive optical networks to tie hundreds of thousands to millions of GPUs together so they can function as a single, planet-scale supercomputer. The ultimate chokepoint for these super-factories isn't just compute, but the network itself. Tying millions of GPUs together requires an optical "central nervous system". To prevent these dense clusters from melting down the power grid, hyperscalers are aggressively transitioning to Co-Packaged Optics (CPO) and silicon photonics to slash energy consumption and eliminate latency bottlenecks. Here is the alpha on who stands to benefit from this massive infrastructure buildout: $NVDA: Pioneering the AI factory architecture with their Spectrum-X and Quantum-X silicon photonics switches, which shatter the limitations of legacy networks. $SOI: The hidden monopoly at the base of the supply chain, commanding >95% market share for the Photonics-SOI engineered substrates required to fabricate these optical chips. Crucially, the transition to CPO drives a massive 4x wafer content multiplier per optical port, supercharging their growth. $LITE & $COHR: The indispensable suppliers providing the high-power lasers and Optical Circuit Switches (OCS) required to light up and dynamically route these massive data fabrics. $MRVL: The "Switzerland" of AI connectivity, providing the critical optical DSPs and scale-up "Photonic Fabric" needed to break the memory wall inside the rack. $FN: The premier contract manufacturer handling the complex precision assembly and packaging of 1.6T transceivers and CPO components for hyperscale networks. $VRT & $CEG: The physical backbone of the factory, providing the essential liquid cooling systems and gigawatt-scale carbon-free nuclear power required to keep these thermodynamic heavyweights running. $NVDA $SOI $LITE $COHR $MRVL $FN $VRT $CEG

Morgan Stanley just upgraded Soitec $SOI $SLOIF from equalweight to overweight with a fresh price target of €70 from €35 This is 28% higher than the current market price. The stock is up slightly in European markets. As silicon photonics scales from pluggables toward co-packaged optics, Soitec doesn’t just benefit from volume, it also enjoys a positive mix shift that should meaningfully expand gross margins, a dynamic Morgan Stanley believes is still underappreciated in the shares.

@PhotonBull #soitec

@TheStoicCapital $SOI #soitec

$VIAV is the most mispriced stock in the photonics super-cycle, but I do not see people speaking about it? $COHR — 251x earnings. Growing 23% a year. $VIAV — 25x earnings. Growing 45% a year. VIAV is growing FASTER. At a TENTH of the valuation. —————— What does Viavi actually do? Every transceiver $LITE and $COHR ship gets tested on Viavi equipment before it leaves the building. No exceptions. No alternatives. It’s mandatory. They are the toll booth on the entire photonics supply chain. As the photonics cycle plays out, so will VIAV. —————— The numbers: Revenue growth: 6% → 36% in 4 consecutive quarters Photonics segment alone: +45% YoY Operating margins: expanding from 14.9% → 19.3% EPS growth: +46% YoY This is an accelerating compounder nobody’s priced correctly. —————— The bear case? Not a pure play. 21% of revenue is banknote authentication. Nothing to do with AI. Strip that out entirely, the photonics division is growing 45% and you’re basically getting it free at this valuation. —————— Everyone bought the laser companies. Nobody bought the company that certifies every laser works. $VIAV still think there’s a re rating to happen here👀