Investment Research · Published March 16, 2026 · Updated July 9, 2026

The Future of Compute
is Supply-Constrained.

Eleven research reports spanning semiconductor bottlenecks, robotics, inference scaling, cross-border equity signals, and AI capital-allocation maps.

SemiAnalysis × DwarkeshIEA · ASML · FERC50 X/Twitter stocks tracked
Scroll to explore
0
EUV tools shipped
ASML, 2025
0
Companies tracked
Global universe
0%
Median YTD return
AI/GPU buildout
0x
ASML multiplier
$400M → $14.3B
Eleven Research Reports

Choose Your Entry Point

Key Findings Across All Reports
01

The bottleneck shifts every 2–3 years

Memory & packaging (2026–27) → Power (2028) → Fabs & EUV (2029–30) → Geopolitics (2031+). Each phase creates different winners.

Source: IEA, ASML, FERC, SIA/BCG
02

HBM, storage, and packaging have been discovered

June refresh: HBM +237.4% median, storage/data +200.0%, advanced packaging +164.0%. The highest current alpha now shifts toward power, cooling, and selected packaging names that still have bottleneck exposure without the same rerating.

Source: Yahoo Finance, IEA, Jun 2026
03

ASML is the most asymmetric trade

A $400M EUV tool enables $14.3B downstream value. ASML captures <3% of what it creates. Pricing power inflects 2028–2032.

Source: ASML 2025 Annual Report
04

Robotics is escaping the teleop trap

EgoScale: 20,854h of human video with R²=0.9983 scaling law. DreamDojo: 44,711h. That’s 89.4x Figure’s teleop data. The scaling substrate is shifting from robot demos to human video.

Source: NVIDIA EgoScale, DreamDojo (Feb 2026)
05

Structure beats brute-force thinking

Structured test-time scaling — recursion, context isolation, verification — outperforms naive chain-of-thought. Value shifts to verifier infrastructure and recursive training flywheels.

Source: arXiv: RLM, MiroThinker-H1, ATTS
06

880+ assets and episodes mapped across the research stack

100 GPU buildout equities, 100 passives residual-alpha names, 120 unified semiconductor/CPO alpha names, 100 semiconductor AI node/connection names, 100 latent AI company nodes, 50 robotics companies, 100 test-time scaling names, 50 cross-border signal names, and 162 podcast episodes. Each is scored on chokepoint exposure, mispricing, scaling alignment, method fit, network centrality, or hidden alpha.

Source: All reports combined
How to Navigate

Eleven Reports, One Thesis

Start with the Bottleneck

Report I maps the physical constraints — EUV tools, power grids, memory fabs — that limit how fast AI can scale. This is the macro framework.

🏭
Then the Supply Chain

Report II zooms into 100 public companies across 10 sectors. Filter by sector, sort by current alpha, and explore bull/bear theses for each name.

🤖
Then Robotics

Report III asks: can robots learn from human video instead of expensive teleop? 10 methods scored, 5 labs profiled, 50 companies ranked.

🧠
Then Scaling Intelligence

Report IV examines whether AI can get smarter by structuring its own thinking. 5 novel RLM methods, 16 scenarios, 100 companies.