Optimizing the energy consumption of modern ML
GPU-to-Grid: Voltage Regulation via GPU Utilization Control
Where Do the Joules Go? Diagnosing Inference Energy Consumption