Search papers, labs, and topics across Lattice.
This study addresses the challenge of balancing energy consumption and performance in High Performance Computing (HPC) systems by implementing dynamic power regulation strategies for mixed workloads. It compares two feedback control methods: a gain scheduled proportional-integral (PI) controller and a polytopic linear parameter-varying (LPV) controller, both of which adapt to workload changes. The results show that the LPV controller outperforms the PI controller in terms of tracking error, control variance, and transient behavior during workload phase transitions, while adhering to power constraints.
LPV control reduces tracking error and improves stability in dynamic power capping for HPC systems, outperforming traditional methods.
Balancing energy consumption and performance remains a critical challenge in High Performance Computing (HPC) systems. While static power capping mechanisms such as Intel's Running Average Power Limit (RAPL) offer basic control capabilities, they lack the flexibility to adapt to dynamically varying workloads. This work studies dynamic power regulation for mixed workload scenarios. We investigate two feedback strategies: a gain scheduled proportional-integral (PI) controller and a polytopic linear parameter-varying (LPV) controller synthesized via H$\infty$ control, both scheduled by a workload indicator that changes between memory and compute phase. We evaluate tracking performance, phase switching, and robustness under practical power cap constraints. While both controllers respect power limits, the LPV design achieves lower tracking error, lower control variance, and smoother transients during phase changes than gain scheduled PI.
