PIM-CACHE: High-Efficiency Content-Aware Copy for Processing-In-Memory

2026-03-24

AI Summary

This work addresses the performance bottleneck in Processing-in-Memory (PIM) architectures caused by redundant, coarse-grained data transfers between the host and DPUs. To mitigate this inefficiency, the authors propose PIM-CACHE, a lightweight, content-aware data staging layer that introduces content-aware mechanisms into PIM data movement for the first time. By dynamically identifying workload similarity at runtime, PIM-CACHE eliminates redundant data copies and optimizes cache management. Implemented on the UPMEM PIM platform and evaluated using both synthetic and real-world genomic datasets, the approach significantly reduces data transfer overhead and enhances overall system efficiency.

Technology Category

• Machine Learning: Hardware-aware ML
• Data Mining & Knowledge Management: Scalability, Parallel & Distributed Systems
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Data management and stream processing for Web, mobile and wireless applications
• Security and Privacy: Data transparency and provenance
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

Processing-in-memory (PIM) architectures bring computation closer to data, reducing the processor-memory transfer bottleneck in traditional processor-centric designs. Novel hardware solutions, such as UPMEM's in-memory processing technology, achieve this by integrating low-power DRAM processing units (DPUs) into memory DIMMs, enabling massive parallelism and improved memory bandwidth. However, paradoxically, these PIM architectures introduce mandatory coarse-grained data transfers between host DRAM and DPUs, which often become the new bottleneck. We present PIM-CACHE, a lightweight data staging layer that dynamically eliminates redundant data transfers to PIM DPUs by exploiting workload similarity, achieving content-aware copy (CAC). We evaluate PIM-CACHE on both synthetic workloads and real-world genome datasets, demonstrating its effectiveness in reducing PIM data transfer overhead.

Problem

• Processing-in-Memory
• data transfer bottleneck
• coarse-grained data transfers
• PIM architectures
• memory bandwidth

Innovation

• Processing-in-Memory
• Content-Aware Copy
• Data Transfer Optimization
• PIM-CACHE
• Workload Similarity