ReVolt: Power Delivery Network-Aware Voltage Droop Control for 2.5D PIM Chiplet Architectures

2026-08-09

AI Summary

This work addresses voltage droops in 2.5D PIM multi-chiplet architectures caused by chiplet-level current surges, which lead to voltage violations, performance degradation, and loss of ML inference accuracy. To mitigate these issues, the paper introduces ReVolt, a novel framework that, for the first time, treats operational unit (OU) scale as a runtime control variable. ReVolt integrates an LSTM-based power delivery network surrogate model to predict per-chiplet voltage trajectories in real time and dynamically schedules OUs to proactively suppress current demand. This proactive approach effectively alleviates voltage droops without compromising inference accuracy, achieving a 76× average reduction in energy-delay product (EDP) compared to the baseline while significantly improving energy efficiency.

Abstract

Processing-in-memory (PIM)-based 2.5D multi-chiplet platforms are enablers for machine learning (ML) workloads. However, their performance is affected by the power delivery network (PDN), where varying chiplet-level current demand induces spatially and temporally varying voltage droop. These droop events lead to voltage violations, degrades system performance, and impact inference accuracy for ML workloads. In this work, we propose ReVolt, a dynamic operation unit (OU)-based framework for mitigating voltage droop in PIM-based multi-chiplet systems. ReVolt leverages an LSTM-based PDN surrogate to predict per-chiplet supply voltage trajectories at runtime, enabling proactive adjustment of OU size to mitigate droop events. By treating OU size as a control knob, ReVolt regulates chiplet-level current demand while maintaining computational accuracy. This approach prevents voltage droop violations and improves energy-delay product (EDP) while preserving ML model inference accuracy. Experimental results demonstrate that ReVolt prevents voltage droop violations while achieving an average 76x reduction in EDP compared to existing fixed and dynamic OU-based baselines, without compromising inference accuracy of ML models.

Problem

• voltage droop
• power delivery network
• 2.5D chiplet
• processing-in-memory
• machine learning

Innovation

• voltage droop mitigation
• power delivery network (PDN)
• processing-in-memory (PIM)
• dynamic operation unit
• LSTM-based surrogate model