Where each company stands
Listing companies one by one hides the question that matters. Placed on the terrain, the relationships become readable.
The terrainsegment flow / company dependency
Silicon
Foundry
4
KR005930
TWTSM
NLASML
USINTC
GPU & Accelerators
7
USNVDA
USAMD
USAAPL
USGOOGL
USAVGO
USINTC
USSMCI
HBM & Memory
4
KR005930
KR000660
USMU
KR042700
Infrastructure
Power
4
USCEG
USGEV
KR034020
KR267260
Data Centers
5
USEQIX
USDLR
USMETA
USVRT
USSMCI
Networking
1
USAVGO
Cloud
5
USMSFT
USGOOGL
USAMZN
USORCL
USCRWV
Models & Tokens
Foundation Models
8
USMSFT
USGOOGL
USOPENAI
USANTHROPIC
USAMZN
USNVDA
USMETA
KR035420
Token Economy
3
USOPENAI
USMSFT
USANTHROPIC
Applications
Consumer Devices
2
USAAPL
USQCOM
Ads & Recommendation
2
USGOOGL
USMETA
Commerce
3
USAMZN
KRCPNG
KR035420
Physical AI
Robotics
3
KR005380
KR277810
USTSLA
Autonomy
2
USTSLA
KR005380
Solid — flow and supply
Dashed — competition and self-supply
Thick — bottleneck, hard to replace
Hover a card to see its reach. Click for the full lineage. Drag the background to pan.
A moat is not an attribute but a position in the chain. Silicon feeds infrastructure, infrastructure feeds models, models feed applications. Switch to the company view to see who actually supplies whom, and where the single sources are.
Selected segment
Select a segment above to see the companies and the evidence.
Deep filing analysis covers the eight companies with company pages (AAPL–CPNG). All other companies — Korean, Taiwanese, Dutch listings and private AI labs — are placed on the terrain for context only.