Workshops Archive

System Level Energy and Cost Characterization of Hybrid Quantum–Classical Cloud Workloads


Workshop: The 2nd International Workshop for Software Frameworks and Workload Management on Quantum-HPC Ecosystems

Authors: Waylon Luo, Joao Vitor Macambira Donaton, Priyabrata Senapati, and Qiang Guan (Kent State University)

Abstract: Near-term quantum workloads depend on hybrid quantum-classical cloud architectures, yet their power and performance dynamics remain difficult to characterize. To address this, we present a high-level simulation framework to evaluate power, energy, and cost across heterogeneous QPU and CPU clouds. By modeling macro level interactions such as job arrivals, workflow iterations, and device power states, the framework orchestrates administrative-level evaluation without requiring detailed quantum properties. To demonstrate the framework with selected configurations and workload, we sweep the variational iteration count over a fixed fleet, holding circuit width, depth, shots, and arrival patterns constant, so that loop length serves as the sole independent variable. While computation and per-job energy scale linearly, QPU occupancy and turnaround depart abruptly from linearity beyond a critical iteration count: past it, nearly all quantum phase time is spent blocked on qubit connectivity rather than computing, and turnaround per job becomes heavily tailed.


Back to The 2nd International Workshop for Software Frameworks and Workload Management on Quantum-HPC Ecosystems Archive Listing

Back to Full Workshop Archive Listing