Mechanisms · category

Remote & side-channel sensing

Inferring activity from outside or from physical signals: detecting data centres, classifying workloads from power or other emissions.

NameTypeReadinessVerifiesThreat model
On-chip telemetry from timing, memory and performance counters
Uses timing, memory-residency and performance-counter signals measured on AI accelerators as evidence about which workloads they are running.
MechanismR2DemonstratedThis compute runs inference, not training; Declared hardware is idle or shut downSemi-trusted prover
Workload classification from telemetry and side channels
Telling whether chips are training, serving or doing non-AI work from GPU counters or power draw, signals that do not read weights or data.
MechanismR2DemonstratedThis compute runs inference, not trainingAdversarial prover
Remote detection of data centres
Remote detection locates large data centres and estimates their power capacity without site access, using satellite imagery, heat signatures and public records such as permits.
MechanismR1ProposedThere is no undeclared relevant computeAdversarial prover

Includes records that list this as a secondary category.