Skip to content
FoxworthFOXWORTHThe Archive
← ArchiveMonitoring

THE COMPUTE FRONTIER

The accelerator supply chain that decides who can train the largest models.

Opened 18 MAR 2026Last update 12 MAY 20261 sources

Summary

Frontier AI is gated by access to advanced accelerators, the power to run them, and the fabs that make them. This File tracks how compute became a strategic resource priced and rationed like one.

The bottleneck is not ideas but the physical stack beneath them: silicon, memory bandwidth, interconnect, and electricity.

Timeline

  1. 2012

    GPU inflection

    Deep learning demonstrates that general-purpose GPUs vastly outperform CPUs for training, reorienting the industry.

  2. 2023

    Accelerator scarcity

    Demand for leading accelerators outstrips supply; allocation becomes a competitive advantage in itself.

Key Actors

Accelerator vendorscompany

Designers of frontier training hardware

Hyperscalerscompany

Largest buyers and operators of compute

Grid operatorsinstitution

Constraint on where compute can physically grow

Related Files

Signals

SIG-014228 MAY 2026probable

Reported leading-edge fab utilization remains near capacity despite a broad hardware glut elsewhere, suggesting the chokepoint in FILE #001 is tightening, not easing.

SIG-014020 MAY 2026confirmed

Large power transformer lead times are again being quoted in years, reinforcing the binding constraint identified in FILE #003.

SIG-013912 MAY 2026probable

Accelerator allocation is increasingly negotiated as multi-year supply commitments rather than spot purchases — compute behaving like a strategic reserve (FILE #002).

SIG-013621 APR 2026confirmed

Datacenter siting decisions are increasingly driven by grid interconnection availability rather than land or tax — FILE #002 and FILE #003 are converging.

Open Questions

  • ?Does power, not silicon, become the binding constraint on compute within the decade?

Sources

  • Compute and capability scaling — public research · 2023