To address the growing energy demand and environmental issues associated with fossil fuels, growing energy crops on marginal lands offers a sustainable solution to increase bioresources for bioenergy production while avoiding the conflict of food and fuel. Giant reed (Arundo donax L.), a non-food energy crop, is recognized for its substantial yield and adaptability to marginal lands, making it a promising candidate for bioenergy production and carbon sequestration. Despite its potential, the optimal supply chain configuration for giant reed (GR) based bioenergy production on China's marginal lands has not been established and evaluated. This study aims to fill this gap by developing and evaluating an integrated framework that includes the identification of available marginal lands, a region-specific growth model for GR biomass, and a Mixed-Integer Linear Programming (MILP) model for optimal design of spatial supply chain configuration. The growth model indicates an average yield potential of 24.5 t DM/ha annually. The optimization model provides multiple cost-effective solutions for the configurations of biomass production regions, storage depots, biorefineries, while considering transportation logistics across diverse geographic landscapes. Results at various replacement ratio of national energy consumption, demonstrate that large-scale GR plantation and utilization significantly increase the total cost but also contribute to substantial carbon sequestration, ranging from 23 Mt CO2 eq at a 1% replacement ratio to 620 Mt CO2 eq at 28% annually. This study not only provides a foundational framework for future research on the potential of GR in China's bioenergy development, but also highlights the feasibility of incorporating GR into China's bioenergy landscape.