# Benchmark — Handheld (Soldier System) Blueprint For platform-level benchmarking — vision AI inference, hardware-accelerated media processing, end-to-end video analytics pipelines, and generative AI — see the [Edge Workloads and Benchmarks Guide](./benchmarking/run-edge-benchmarks.md). The Handheld Multi-Modal application itself reports live CPU, memory, GPU, NPU, and power telemetry in the **Resources** panel of the operator console. See the [Operator Guide](operator-guide.md#resources). :::{toctree} :hidden: Edge Workloads and Benchmarks Guide <./benchmarking/run-edge-benchmarks.md> :::