Data-center-enabled supply chains: How compute proximity and architecture affect planning latency and inventory efficiency

This paper focuses on how supply chain planning latency and inventory efficiency in mid-market manufacturing in the United States can be influenced by data center architecture, i.e., edge, cloud, and hybrid, and especially on small and medium-sized enterprises (SMEs). With world uncertainties such as pandemics and geopolitical stressors, real-time decision-making in AI-driven supply chains is dependent on proximity to compute, but the architecture may be inflated with latency, inventory volumes, cost-to-serve, carbon emissions, and service failures. Using a mixed-methods design, which combines simulation modeling and stress scenario analysis of anonymized time-series data, we measure these effects: Edge reduces latency by 40-60 percent compared to cloud but also raises local energy consumption; hybrids trade off, reducing safety stock by 20-30 percent and preserving service levels exceeding 95 percent in 50 percent demand spikes. Products are four architectural reference patterns (e.g., edge-dominant in the case of volatile routes), latency to value curves with diminishing returns (e.g., 100 MS latency reduces inventory by 15-22%), and policy advice to the adoption by SMEs. Results indicate that hybrids outperform in terms of cost efficiency ($12/order) and sustainability maximization by incorporating compute infrastructure into the supply chain theory. Architecture is based on practical implications, whereas policy recommendations support subsidies, training, and standards as a means of bridging SME digital gaps. This study will contribute to resilient and efficient Industry 4.0 supply chains.