A

Anthropic

PBC · $380B · Founded 2021

AnalyticsSupply ChainCompetitorsFunding

Supply Chain Control Room

A more visual map of how Anthropic depends on infrastructure, partners, and operational chokepoints.

The stack below turns tier data into an exposure view: where dependencies cluster, which layer carries the most vendors, and how many live risk signals are already mapped.

Mapped tiers

6

Critical Elements

Node clusters

16

24 listed dependencies

Risk signals

18

4 formal assessments

Highest severity

critical

14 flagged nodes

T01Frontier Models
T02Infrastructure
T03Processors
T04Memory & Network
T05Foundries & Machines
T06Critical Elements

Exposure Mix

Operational risk by signal density

Live map
Signals18across 6 tiers
critical4
high5
medium7
low2

Dominant concentration

Critical Elements carries the widest cluster with 4 mapped nodes.

Dependency Lanes

Each tier is rendered as a lane, with grouped nodes and dependency payloads surfaced directly inside the mapped layer.

6 tiers · 16 nodes · 24 dependencies
01

Tier 01

THE MINDS

Frontier Models

Claude model family.

Node clusters

2

Dependencies

7

Node

Claude Models

Items

3

  • Opus 4.6
  • Sonnet 4.6
  • Opus 4.7 (NEW)
Node

Revenue Hubs

Items

4

  • API / Claude.aiDIRECT
  • Amazon Bedrock ~50%PRIMARY
  • Vertex AISECONDARY
  • Azure FoundryNEW
02

Tier 02

THE LANDLORDS

Infrastructure

Multi-cloud, AWS-heavy.

Node clusters

3

Dependencies

4

NodePRIMARYHIGH

Amazon AWS

Items

2

  • Rainier: 500K+ → 1M+ chipsSCALING
  • 5 GW energyPOWER
NodeSECONDARYMEDIUM

Google Cloud

Items

1

  • TPU v6: Largest non-Google userUNIQUE
NodeBURSTLOW

CoreWeave

Items

1

  • B200 burstNICHE
03

Tier 03

THE ORGANS

Processors

Silicon Agnostic — 3 architectures.

Node clusters

3

Dependencies

3

NodeMEDIUM

Trainium 3/2

Items

1

  • Annapurna Labs customCUSTOM
NodeMEDIUM

TPU v6

Items

1

  • Training + inference
NodeHIGH

Nvidia B200 & Rubin

Items

1

  • Via cloud partnersINDIRECT
04

Tier 04

NERVOUS SYSTEM

Memory & Network

#1 bottleneck.

Node clusters

2

Dependencies

4

NodeHIGH

HBM4 Memory

Items

3

  • SK Hynix 60%
  • Samsung
  • Micron 24%
NodeMEDIUM

Broadcom

Items

1

  • Tomahawk 5 switchesCRITICAL
05

Tier 05

THE FACTORY

Foundries & Machines

Every AI chip comes from these two companies.

Node clusters

2

Dependencies

2

NodeCRITICAL

TSMC

Items

1

  • Only 3nm/2nm foundryMONOPOLY
NodeCRITICAL

ASML

Items

1

  • High-NA EUV. No alternative.NO ALT
06

Tier 06

THE INGREDIENTS

Critical Elements

The Silicon Curtain of 2026.

Node clusters

4

Dependencies

4

NodeCRITICAL

Gallium & Germanium

Items

1

  • 25% revenue tax on chip salesGEOPOLITICAL
NodeCRITICAL

Neon Gas

Items

1

  • ~50% of semiconductor-grade neonWAR ZONE
NodeHIGH

Copper

Items

1

  • Record prices hit DC costsCOST
NodeMEDIUM

Silicon Wafers

Items

1

  • Shin-Etsu & SUMCODUOPOLY

Risk Assessments

Formal assessments are separated from node-level flags so the page reads like a dashboard first and a report second.

4 assessments identified

Severity Stack

critical
4
high
5
medium
7
low
2

Immediate read

The mapped risk posture currently peaks at critical severity, with concentration strongest in Critical Elements.

high

Compute Dependence

Training on infra owned by Amazon, Google, Microsoft — same companies with competing models. If any relationship sours, model quality falls within months.

low

Pentagon Clash

Publicly clashed with DoD over military AI use. May limit government contract revenue.

medium

Revenue Accounting Dispute

OpenAI alleges Anthropic gross-up of cloud partner revenue inflates ARR by ~$8B. Comparable figure may be ~$22B vs headline $30B.

medium

IPO Timing Pressure

Preparing IPO potentially in 2026. Hired Wilson Sonsini. Growth rate must sustain through public market scrutiny.