Mechanism · Workload classification from telemetry and side channelsLimited scale and hardware coverageOverviewEvidence & limitsTechnical detailSourcesOn this pageFindingSources← All known flawsSignificantOpen questionOpenRahman and Tajdari's experiments are single-node, with up to 8 GPUs, and cover NVIDIA hardware only 1. Gargiulo and Kulp's corpus was collected on a single H200 NVL, and generalization across devices remains to be established 2. Sources: [1] · [2]