Would you rather suck on my nipples or my clit?
Top-5 semicap DRAM equipment revenue up 80% y/y in 3QCY26. Another ATH and ASML gaining share. $35B+ annual run-rate for top-5 semicap DRAM equipment revenue in 3Q26. Showing no signs of slow down. KLA, for example, expects 60% of its process control revenue to come from memory (mostly DRAM), up from 30%+ in 2026.
Show more
Intel's server CPU volume up 10% y/y while ASP up 54% in 3Q26. Units are still far below the previous peak but stabilizing. From 2H25 Intel started the effort of triaging internal supply for DC and the effort began to pay off from early this year (ASPs up strong double-digits y/y for 3 consecutive quarters). Intel even going to the lengths of reducing 18A capacity in Oregon to deploy Intel 3, which is the work horse node for server CPUs.
Show more
Kioxia QSep26
- Revenue up 39% q/q and 405% y/y to $15.4B (above the $14.8B guide) at 84% GM; QDec guided to $15.9B
- Blended ASP up mid-20s% q/q, while bit shipments up low-teens
- CY27 capacity fully sold out; LTAs cover 50%+ of CY28 bit shipments
- NAND bit demand growth high-teens in CY26 and low-20s in CY27
- DC and enterprise 65% of SSD
Show more
STMicro 3Q26
- Accelerating recovery & AI DC story; AI DC revenue now tracking $1.3B in 2026 and well above $2B in 2027; total book-to-bill well >2 (and >3 in industrial and comms)
- Revenue up 10.5% q/q & 24.7% y/y to $3.76B at 37.3% GM (+250bps q/q); QDec guide $4.15B (10.5% q/q) at 39% GM
- CY26 revenue tracking $14.5B (+22.8% y/y); below the $17.3B FY23 peak
- SiC will grow 20% in 2026 (driven by European EV)
- SiPho capacity to grow >4x y/y for 2027
- Inventory stable at 3.25B with 126 days DOI (flat q/q)
Show more
ASE, the packaging powerhouse, revised its capex for 3rd time this year to $11.5B. 2027 capex to be more vs 2026. ASE capex > all non-China mature foundries. Also raised its leading-edge advanced packaging guidance for 2026 by 6% and expects to double in 2027.
Show more
Welcome, Startup Fellowship Cohort 003, and thank you to all who applied.
Seoul, Day 1 is in the books.
Thank you to everyone that came out yesterday.
Day 2 // 10:00-20:00 KST
744-19 Hannam-dong, Yongsan-gu
Stay sovereign.
THE SOVEREIGN AI REVOLUTION IS IN FULL FORCE.
We are helping our customers build superintelligence they own, with data they control.
McCarthy Building fine-tuned Nemotron in Foundry on labeled RFI data in the Ontology to identify project risks before they escalate, outperforming frontier models at significantly lower cost.
Tampa General is using sovereign AI to better protect patient data, keeping sensitive data and model processing local, and fine-tuning Nemotron models to classify LLM traffic against its own risk and privacy standards.
NVIDIA is using Palantir AIP and Nemotron to accelerate its own supply chain, shortening the path from wafer to deployment, spotting disruptions before they hit production, and keeping its models under its control.
L3Harris is bringing AI workflows in-house and building its own models trained on proprietary data to support critical defense use cases and the warfighter.
Ondas is launching Regna AI on Palantir AIP, leveraging open-weight models including NVIDIA Nemotron, to help customers adapt, evaluate, and deploy mission-specialized AI while retaining control of their data, models, and infrastructure.
Show more
TSMC predicts A14 to be a bigger node than N2. Every new node is now breaking records. TSMC's N2 node grew 3x q/q to $3B+ in 3Q26. The node also happens to be the first billion dollar opener for TSMC nodes when it hit $1B+ in 2Q26. That $1.0B was just 3% of wafer revenue vs 11% for N7 at its debut as TSMC's quarterly wafer base grew >4x during this time. Every generation now starts smaller in share and larger in dollars. N2's 14% jump over N3 is the biggest generational step in a decade.
Show more