Reproduce “SoK: Systematizing a Decade of Architectural RowHammer Defenses Through the Lens of Streaming Algorithms” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Michael Jaemin Kim, Seungmin Baek, Jumin Kim, Hwayong Nam, Nam Sung Kim, Jung Ho Ahn
Researchers from Seoul National University, UIUC, and Meta systematize ten years of architectural RowHammer defenses, re-framing them using streaming algorithms to create a unified taxonomy and quantify hardware overheads. The analysis identifies vulnerabilities in prior empirical defenses and proposes theoretically robust alternatives with mathematical security guarantees against evolving attack patterns.
Key insights
- Many existing architectural RH defenses implicitly or explicitly "reinvent" established streaming algorithms, underscoring the value of a formal algorithmic framework for understanding and developing solutions.
- The choice and implementation of streaming algorithms for trackers significantly impact hardware overheads (area, capacity) and their suitability for different mitigative actions (e.g., refreshing rows requires careful counter decrements).
- Empirically validated sampling-augmented tracker-based defenses can still be vulnerable to sophisticated, phase-changing adversarial patterns, necessitating solutions with mathematically proven worst-case security guarantees.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.