A comparative study on power delivery aspects of compute-in/near-memory approaches using DRAM

2026-04-06

AI Summary

This work addresses the unique power delivery challenges introduced by DRAM-based processing-in-memory (PIM) architectures, which exhibit unconventional current demand patterns that exacerbate voltage droops, IR drops, and thermal hotspots in the power delivery network (PDN). The paper presents the first PDN-aware current behavior classification framework tailored for DRAM-PIM, systematically characterizing the spatiotemporal current signatures induced by diverse PIM operations—including multi-row activation, row-buffer computing, and near-bank compute units. By integrating timing constraints, memory controller scheduling, data layout strategies, and bank-level power management, the study elucidates the underlying mechanisms through which different computational paradigms impact PDN stability. The findings underscore the critical importance of PDN-conscious design in enabling scalable and reliable DRAM-PIM systems and provide clear directions for future optimization.

Abstract

Compute-in-memory (PIM) mitigates the memory wall by performing computation within memory, reducing data movement and improving energy efficiency. DRAM-based PIM is particularly attractive due to its high density, mature manufacturing ecosystem, and compatibility with existing systems. Recent works exploit multiple levels of the DRAM hierarchy - including subarrays, banks, and 3D-stacked organizations - to enable in-memory computation using mechanisms such as multi-row activation, row-buffer operations, and near-bank compute units. However, these approaches introduce non-traditional current demand patterns that challenge the power delivery network (PDN). This paper surveys PDN challenges in DRAM-based PIM systems and proposes a unified taxonomy that characterizes PIM-induced current behavior along temporal (burst vs. sustained) and spatial (localized vs. distributed) dimensions. Using this framework, we analyze how representative PIM techniques stress the PDN through bursty activations, multi-row concurrency, and large-scale parallel execution, leading to voltage droop, IR drop, and thermal hotspots. We further discuss DRAM-specific mitigation strategies leveraging existing architectural and circuit-level mechanisms, including timing constraints, memory controller scheduling, data placement, and bank- and vault-level power management. This survey highlights the importance of PDN-aware design for scalable and reliable DRAM-based PIM systems and outlines key future research directions.

Problem

• Compute-in-Memory
• DRAM
• Power Delivery Network
• Voltage Droop
• IR Drop

Innovation

• Compute-in-Memory
• DRAM
• Power Delivery Network
• Voltage Droop
• In-Memory Computing