PVAC: A RowHammer Mitigation Architecture Exploiting Per-victim-row Counting

2026-04-22

AI Summary

This work addresses a critical limitation in DDR5’s Probabilistic RowHammer Attack Counter (PRAC) mechanism, which counts all row activations—including benign refreshes—and relies on explicit RFM resets, often leading to counter saturation, false mitigation triggers, and unnecessary performance degradation. To overcome this, the authors propose PVAC (Probabilistic Victim-Aware Counter), a novel approach that aligns counter semantics with the physical disturbance model of RowHammer. By design, PVAC naturally bounds counter values under normal refresh operations and eliminates interference from irrelevant activations. Implemented via a dedicated DRAM Counter Sub-Array (CSA), PVAC enables concurrent counter updates without timing overhead and incorporates a low-refresh-cost layout optimization. Evaluations demonstrate that PVAC completely eliminates false positives under both benign workloads and adversarial RowHammer attacks, substantially improving hammering tolerance while simultaneously enhancing security, performance, and energy efficiency.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Hardware-aware ML
• Search and Optimization: Distributed Search

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Security and Privacy: Tracking, profiling, and countermeasures against them
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

As DRAM scaling exacerbates RowHammer, DDR5 introduces per-row activation counting (PRAC) to track aggressor activity. However, PRAC indiscriminately increments counters on every activation -- including benign refreshes -- while relying solely on explicit RFM operations for resets. Consequently, counters saturate even in an idle bank, triggering cascading mitigations and degrading performance. This vulnerability arises from a fundamental mismatch: PRAC tracks the aggressor but aims to protect the victim. We present Per-Victim-row hAmmered Counting (PVAC), a victim-based counting mechanism that aligns the counter semantics with the physical disturbance mechanism of RowHammer. PVAC increments the counters of victim rows, resets the activated row, and naturally bounds counter values under normal refresh. To enable efficient victim-based updates, PVAC employs a dedicated counter subarray (CSA) that performs all counter resets and increments concurrently with normal accesses, without timing overhead. We further devise an energy-efficient CSA layout that minimizes refresh-induced counter accesses. Through victim-based counting, PVAC supports higher hammering tolerance than PRAC while maintaining the same worst-case safety guarantee. Across benign workloads and adversarial attack patterns, PVAC avoids spurious Alerts, eliminates PRAC timing penalties, and achieves higher performance and lower energy consumption than prior PRAC-based defenses.

Problem

• RowHammer
• PRAC
• victim-row
• DRAM security
• activation counting

Innovation

• RowHammer mitigation
• victim-based counting
• per-row activation counting
• counter subarray
• DRAM security