Papers for
hardware testing teams
Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.
HARE speeds up detecting read disturbance faults in DRAM memory
HARE: Fast Read Disturbance Threshold Profiling in Modern DRAM Chips via Reduced Timing Constraints
Abstract: Modern DRAM chips are vulnerable to read disturbance, which manifests as bitflips in a victim row physically near an aggressor row that is repeatedly activated. State-of-the-art mitigations rely on accurate and exhaustive profiling of the read disturbance threshold (RDT), i.e., the minimum number of aggressor row activations needed to induce the first bitflip. However, RDT changes significantly and unpredictably over time, motivating online profilers that dynamically adjust a mitigation's RDT configuration. RDT profiling is costly for online profilers (e.g., hundreds of seconds for a 16-bank DDR4 module) because 1) measuring a single DRAM row's RDT requires many activations and 2) each DRAM row can have a different RDT, requiring many rows to be profiled. We introduce HARE, an online RDT profiling technique that intentionally exacerbates read disturbance to quickly profile a DRAM module. HARE's key idea is to reduce the timing constraints between consecutive DRAM commands during profiling. Doing so significantly reduces profiling time by 1) directly reducing the execution time of the profiling command sequence and 2) exacerbating read disturbance such that the first bitflip occurs at a smaller activation count than with the conventional RowHammer sequence. From our characterization of 180 real DDR4 DRAM chips from four major DRAM manufacturers, we make three observations. First, HARE measures a DRAM row's RDT 31.2% faster than the RowHammer profiler and a DRAM module's minimum RDT with an average reduction of 21.7%. Second, HARE flips more than 95% of the cells that the RowHammer pattern flips. Third, HARE preserves the RDT ranking of rows well enough such that it can exhaustively reprofile a small set of rows to improve its minimum RDT estimation. Our end-to-end evaluations show that HARE enables significant system performance improvements and outperforms the RowHammer profiler.
Language models improve analog circuit layout repairs with cost-aware planning
Who Plans, Who Edits? Verified, Cost-Aware Division of Labor for LLM-Driven Analog Layout Repair
Abstract: Language-model agents can repair analog layouts when sign-off tools check every edit, but which model should plan a repair, which should edit it, and at what cost remain open questions. In our tool-grounded repair agent, models only choose bounded layout operators; sign-off tools rerun each task's contract checks after every edit, so failing layouts are never accepted. On layout-versus-schematic (LVS), design-rule checking (DRC), and post-layout repairs of three designs (a production high-voltage digital-to-analog converter (DAC), a 5-V oscillator, and a SKY130 operational transconductance amplifier (OTA)) with six tier-matched models from two vendors, 182 blind runs show that cheap models execute plans but rarely succeed alone (the cheapest model: 1 of 11 runs alone, 16 of 17 under a flagship planner; p < 0.001), cheap planners fail more often, and no fixed assignment is cheapest on every task. We propose a learning controller that selects models for each role and the escalation order from the first tool evaluation and updates Bayesian estimates after every attempt. With rules and initial estimates frozen before testing, the controller fixed 7 of 8 episodes on four sealed held-out tasks, matching the OpenAI flagship alone, at 15-40% of the flagship's spend per fix on three tasks but 8% higher spend overall. The cheapest mid-tier model alone fixed 6 of 8 at the lowest overall spend per fix but missed the hardest task. All 32 extracted DAC repairs match the human-guided LVS repair's integral nonlinearity (INL) within 0.005 least significant bits (LSB) at every code.
An efficient one-pass method identifies reliable good chips quickly
Backtracking Candidate Elimination: A One-Pass Algorithm for the Chip Testing Problem
Abstract: In the chip testing problem, we are given $n$ chips, strictly more than half of which are good. Chips can test one another in pairs; a good chip always reports the status of the other chip correctly, whereas a bad chip may report arbitrarily and adversarially. The goal is to identify a single chip that is guaranteed to be good. The problem originates in system-level fault diagnosis and is closely related to the "knights and spies" puzzle. The standard textbook solution is a halving recursion that tests disjoint pairs in rounds and keeps one chip from each consistent pair. We present the Backtracking Candidate Elimination (BCE) algorithm, a sequential alternative that scans the chips once while maintaining a current candidate and a stack of retained chips. Every chip is tested at most once as the incoming chip; when a test is inconclusive the candidate and the incoming chip are discarded together, and the algorithm backtracks to the most recently retained chip. BCE uses at most $n-1$ tests and $O(n)$ time, needs no parity case analysis, and works online. Its correctness follows from two invariants: the retained chips all have the same type, and every discarded pair contains at least one bad chip. We explain how BCE can be viewed as the Boyer-Moore majority vote algorithm with its counter replaced by a stack of physical witnesses, and why that replacement is needed. We also give an early termination rule and a variant for the weaker model of one-directional tests.