A Resilient Edge Inference Architecture: Distributed GPU Compute Beyond the Datacenter
Abstract
This document outlines an alternative architecture for AI inference infrastructure designed to improve resilience, geographic dispersion, and capacity elasticity beyond traditional centralized datacenter models. The proposed system leverages heterogeneous consumer and workstation GPUs as a distributed...
