Energy-aware and Sustainable Cloud Computing: Dynamic Scheduling, Pricing-Driven Resource Management, and Renewable Intermittency

Abstract

Data centers are now the largest users of electricity, carbon, and water, driving rapid cloud computing growth and a global shift to renewable-powered, sustainable cloud infrastructure. The large-scale integration of solar and wind energy creates fundamental operational challenges due to their intermittency, uncertainty, and non-dispatchability. This paper presents dynamic energy-aware cloud computing as a pathway toward sustainability by modeling renewable variability with service reliability assurance aligned workload scheduling with forecast green energy availability. Recent advances in energy-and carbon-aware orchestration (2023-2025) have been reviewed where critically missing gaps are unified frameworks integrating renewable forecasting with stochastic robust optimization SLA aware scheduling plus demand-and supply-side contingencies like storage backup provisioning plus adaptive load management. Based on these gaps an elawful operations model for the co-optimization of power use carbon footprints quality of service under time-varying energy conditions is described that uses elastic resource management supported by probabilistic solar/wind forecasts plus risk-sensitive scheduling. New directions for more sustainable clouds include decentralized speculative renewables infrastructures; this paper also describes some concepts in that area. This work lays out a structured foundation toward next-generation green cloud platforms moving beyond static efficiency metrics into predictive reliability-driven environmentally responsive cloud operations.