Global Chip Supply & Energy Logistics

Overview

Analyzes the physical and geopolitical constraints on AI development, focusing on semiconductor supply chains, manufacturing capacity, and energy infrastructure. It provides a fundamental, ground-level view of what's possible in the world of AI.

What It Does

This pillar aggregates and synthesizes data on three core areas: semiconductor manufacturing, geopolitical trade policy, and energy grid capacity. It tracks lead times for critical AI chips, monitors foundry output from key players like TSMC, and analyzes new government export controls. The pillar combines these signals to assess the real-world bottlenecks affecting AI's growth trajectory.

Why It Matters

AI progress is fundamentally limited by hardware availability and the power to run it. This pillar offers a crucial reality check, identifying supply-side constraints that can delay product launches, cap growth, and impact company valuations long before they are reflected in earnings reports.

How It Works

The analysis begins by tracking monthly production reports from major semiconductor foundries. It then cross-references this with reported order backlogs and lead times for high-demand GPUs from industry sources. Finally, it layers on an analysis of government actions and energy utility filings to model potential chokepoints in the system.

Methodology

A weighted index based on: 1) Chip Availability, measured by weekly tracked lead times for NVIDIA Blackwell and H-series GPUs; 2) Foundry Capacity, using TSMC and Samsung monthly reports to model wafer output; and 3) Constraint Factors, an event-driven score based on new export controls or data center power contract approvals over 100MW.

Edge & Advantage

This pillar provides an edge by focusing on lagging, physical infrastructure indicators that are often ignored in software-centric AI market analysis.

Key Indicators

• GPU Lead Times

  high

  The estimated time from order to delivery for high-end GPUs like NVIDIA's Blackwell series, a direct measure of supply versus demand.

• Data Center Power Contracts

  high

  Approval and size in megawatts of new power agreements for hyperscale data centers, a leading indicator of future compute capacity.

• Export Control Changes

  medium

  Modifications to the US Department of Commerce Entity List or new export restrictions, signaling geopolitical risk to the supply chain.

Data Sources

• TSMC Monthly Revenue Reports

  Provides data on foundry utilization and leading-edge technology demand.

• U.S. Bureau of Industry and Security

  Official source for U.S. export control announcements and Entity List updates.

• Public Utility Commission Filings

  Documents detailing power infrastructure requests and approvals for large industrial customers, including data centers.

• DigiTimes Asia

  Industry publication providing reporting on semiconductor supply chain news and component lead times.

Example Questions This Pillar Answers

• → Will TSMC's 2nm process be in high-volume production by Q4 2025?
• → Will the US implement further restrictions on AI chip exports to any G7 country in 2024?
• → Will the reported lead time for NVIDIA's B200 GPU exceed 52 weeks at any point before 2026?

Tags