Fat Shot Drug

.@SemiAnalysis_ with the great & extremely detailed piece on the 800v transition. Four phases. -Phases 1–2 (2026–28) retrofit existing facilities with a row-level power rack sidecar - no architectural change, just a new box doing AC-to-DC conversion at the rack. -The capex step-up in Phase 1 (~$0.5M/MW) reverses in Phase 2 when the central UPS exits (~$1.2M/MW out). -Phase 3 (late 2028–29) is a real inflection: 800VDC becomes the facility's electrical backbone, AC switchboards and PDUs are eliminated, and a centralized rectifier feeds DC busway directly. -Phase 4 (>2029) is the end state: SSTs replace the LV transformer and rectifier with a single device converting from medium voltage directly to 800VDC, adding ~$0.25–0.75M/MW net but collapsing two conversion stages. Counterintuitive punchline: Throughout all four phases, total electrical capex per MW stays in a $3.6–4.8M band. What changes is who gets paid: central UPS, AC PDU, and LV transformer OEMs are progressively displaced >Winners: SST vendors, SiC suppliers, and DC busway makers capture the incremental spend. Efficiency gains can be up to ~75MW freed per 1GW of IT load (7.5%) by Phase 4. In a dramatically constrained market, every MW counts.

$APGE and their longs have spent years saying that Dupi/Lebri were underdosed and that higher dosing was going to improve efficacy. Today their data show the inverse of that, with high dose underperforming mid dose in every way. Watch how fast their management team completely abandons what (until today) was their key narrative now that it no longer suits them..."Ignore the 1,000 times we said that exposure thing...this thesis is intact and better than ever!" This management team's constant spin bullshit drives me crazy (like saying ISRs are the biggest problem for Dupi - objectively and easily falsifiable bullshit. Only was that was not a *lie* is if you accept they're unintelligent enough to misunderstand what an ISR is. Which is worse?). Anyway, look at their long-awaited high dose P2b results, and compare those to Lebri's (Ebglyss) P2b results...Sorry $APGE, I'm not impressed.

@RollingStone @danawhite We’re just lying now?

The Rolling Stone Interview: Dana White @danawhite tells Rolling Stone he does not speak politically. "You never hear me talk politically, you never hear me say anything left, right. I'm a right down the middle common sense ... that's me."

Just in: Infineon announces price hikes for power semiconductors.

$COGT JP continue to view bezuclastinib as a highly synergistic asset. •mat current valuation levels, $COGT shares significantly discount the strength of the known data for bezuclastinib across the spectrum of indications (2L GIST, NonAdvSM, AdvSM). see a long- term buying opportunity.

$smtc exceeded my already high expectations. Beat, raise, strong H2 commentary, visibility to FY28, and added to list of developments to track.

@LifeOfBoch Cech and Courtois are too low

🚨: Pyramid of the Greatest Goalkeepers of all time. Thought?

往死里炒power semi & MLCCs😁 挑了一只P/B低，产能扩张快、电容、电压、ESL、可靠性高的MLCC股🤔 谢谢大家

chatgpt.com/s/t_6a13b27a01… My gpt asian mlcc cheat sheet

Interesting breakdown by @aleabitoreddit on CPO "substantive monopolies." The risk assessment regarding $FOCI is spot on - their position in passive devices and FAUs faces a severe long-term threat from vertical integration by foundries. However, grouping $HIMX into that same high-risk basket based on a TSMC/Visera threat misreads the underlying hardware physics. Here is why the Visera vs. Himax comparison is fundamentally flawed, and why Himax is a tactical long for Phase 1&2 of CPO: ➡️Incompatible Manufacturing Stacks: While both produce Micro Lens Arrays (MLA), they serve entirely different applications. Visera uses semiconductor lithography designed for image sensors (CMOS). Himax specializes in Wafer-Level Optics (WLO) - literally replicating 3D aspheric structures using specialized optical polymers on glass. They are structurally and materially different components, not 1:1 drop-in substitutes. ➡️The High-Power Optical Barrier: CPO requires MLAs that can handle extreme optical power densities from external laser sources (ELSAM) without thermal degradation. Himax’s proprietary polymer chemistry has spent years clearing these reliability hurdles. Visera’s legacy is in low-power, ambient light sensing. ➡️The "Here and Now" Reality: Is long-term vertical integration a risk with TSMC? Yes, eventually they may want to swallow the entire optical packaging stack. But CPO Phase 1 is commercializing today, and current architectures are already frozen. Re-engineering the pitch-matching and coupling optics to displace Himax would trigger massive requalification delays. ⬇️Bottom line: $FOCI is highly vulnerable near-term as foundries absorb simpler passive paths. Himax, however, protects its near-term upside with highly specialized material science. Even if TSMC attempts to integrate the stack in later cycles, Himax is positioned to print cash during the immediate CPO rollout. 👇 What's your take on the CPO hardware layer? Let's discuss below.

Just in: SK Hynix’s market cap has surpassed $1 trillion.

TSMC plans to hike 3nm prices as much as 15% in the 2nd half 2026 due to surging demand for AI accelerators, media report, adding additional hikes of 5%-10% may follow in 2027. The report cites unnamed supply chain sources. $TSM $NVDA $AVGO $GOOGL $AMZN #semiconductors ctee.com.tw/news/202605277…

Niggas really voted for this LMAOOO

Just heard a wild rumor from someone I trust that works at JPM. The buzz is that Trump has been exploring ways to force Vice President Vance out of office. He cannot fire him legally. Unless Vance resigns, the only way for him to be removed is to be impeached by the House with 2/3rds Senate voting to remove!

Rumblings on the street is that Vance got caught leaking something Trump didn't like, something only Vance could have known.

SK Hynix Effectively Rebuffs US Big Tech's Offers of Tens of Billions of Dollars in Investment Support US big tech companies—Google parent Alphabet, Microsoft (MS), and Meta among them—are pouring out offers to help fund the construction of fabs (chip plants) worth tens of trillions of won and to cover the cost of purchasing extreme ultraviolet (EUV) lithography equipment. But SK Hynix is politely declining these offers while simultaneously using them as leverage in long-term supply agreements. The reasoning: with investment funds already overflowing, there is no reason to disturb the "super supplier" (super-eul) status it currently enjoys. According to industry sources on the 27th, SK Hynix recently received memory-chip capital-investment support offers from Alphabet, MS, and Meta, but management is reportedly not treating them as realistic, actionable options. The biggest sticking point is that building a fab with funding from a specific big tech company would create a contract structure carrying exclusive supply obligations. A semiconductor industry official said, "If you build a production line with a particular customer's money, you create the risk that even if that customer's demand falls in a future downturn, you'd have to prioritize their volume or supply below market price," adding, "Internally, SK Hynix views the side effects that could arise from taking big tech investment negatively and is wary of them." SK Hynix currently holds what is effectively a duopoly alongside Samsung Electronics in the high-bandwidth memory (HBM) market. It supplies most of the HBM mounted on Nvidia graphics processing units (GPUs), and this year's output is already sold out. Because there are virtually no producers other than SK Hynix and Samsung capable of properly mass-producing the high-spec HBM used in AI data centers, a reversal has taken shape in which the big tech customers have instead become the subordinate party (eul) before SK Hynix. The big tech proposals take two main forms. The first is to directly shoulder part of the cost of building the first fab (Y1) in the semiconductor cluster SK Hynix is developing in Yongin, Gyeonggi Province. SK Hynix plans to invest a total of 31 trillion won in the first fab alone; once completed, it will add 350,000 wafers per month of production, expanding total capacity to around 900,000 wafers per month. Also under discussion was a plan to subsidize the cost of purchasing EUV lithography equipment from the Netherlands' ASML. ASML's latest High-NA EUV machine is an ultra-premium tool costing roughly $400 million (about 550 billion won) per unit—about twice as expensive as the current-generation EUV. To secure this equipment, which is essential for mass-producing advanced memory at the 1c DRAM node and beyond, SK Hynix has already finalized a plan to bring in some 12 trillion won worth of EUV tools. The move is read as big tech's intent to effectively pre-secure dedicated production lines by sharing part of this cost. Behind the unusual nature of these proposals lies the extreme intensification of the AI infrastructure investment race. Alphabet, Meta, and MS each disclosed AI infrastructure investment plans worth tens of billions of dollars in their recent earnings reports. MS in particular said its capital expenditure this year would reach $190 billion, explaining that the increase in component costs such as chips alone amounts to $25 billion. Meta likewise acknowledged its supply-chain pre-emption strategy, saying it is signing contracts across the supply chain to secure necessary components in advance. SK Hynix is refusing structural dependence—such as fab equity investment or joint equipment ownership—while opting instead to use big tech's desperation as negotiating leverage. A source familiar with SK Hynix said, "Various ideas are being discussed to strengthen the binding force of contracts as we sign long-term agreements (LTAs) with global big tech." The industry interprets this as higher prepayments, ultra-long contracts of five years or more, and price-floor guarantee clauses. The strategy is to convert the funds big tech brings not into factory equity, but into more favorable contract terms. SK Hynix's confidence is backed by the numbers. Its operating margin as of the first quarter this year reached 72%, and in February it announced an additional $15 billion (about 21 trillion won) investment plan to expand next-generation memory production. Construction of the first fab at the Yongin cluster is under way, targeting its first cleanroom operation in February next year—three months ahead of the original schedule. $MU $DRAM

@MM3915038575333 Agreed

@Biotech_FC $CYTK should be trading at $100..

I feel like biotechs are hated right now lol

Example: CPO testing or HBM - bottleneck huge future demand - power semis & MLCCs Supply crunch - raw materials for CCL