University of Thessaly (UTH)
Advanced electromigration analysis methods for long - term reliability and hardware security in complex VLSI interconnects
Abstract
dc:descriptionThe continuous scaling of integrated circuits (ICs) has introduced significant reliability and security concerns, especially in dense VLSI power grid interconnects. Among these concerns, electromigration (EM) has emerged as a dominant failure mechanism, capable of degrading the performance and lifetime of modern chips. It is a physics-driven effect, where momentum transfer from conducting electrons to metal atoms induces atomic migration along interconnect wires. This process causes gradual material displacement under high current densities, resulting in void or hillock formation that may eventually lead to open or short circuits. As technology nodes shrink and current densities rise, EM analysis becomes increasingly critical to ensuring long-term IC reliability. In practice, conventional EM assessment methods are primarily empirical and fail to capture the complex spatiotemporal behavior of EM in modern multi-segment interconnect structures. While recent physics-based models offer greater accuracy, most implementations are limited to either the pre-void phase or rely on steady-state assumptions that become invalid under realistic operating conditions. Moreover, emerging hardware threats have revealed that EM behavior can be maliciously manipulated to introduce undetectable hardware Trojans, exposing a security blind spot in traditional reliability methodologies. To overcome these limitations, this dissertation introduces scalable, physics-based methods for EM analysis, with attention to both reliability and security aspects in VLSI interconnects. It first proposes a complete and scalable EM analysis framework that captures both the pre- and post-void phases by employing matrix exponential formulation combined with Krylov-based techniques. Then, a moment matching (MM) approach using asymmetric Krylov projections is introduced to accelerate EM simulations on large-scale interconnect structures. Finally, the work presents a security-aware EM modeling strategy that introduces design-level countermeasures to safeguard power grids and detects malicious structural changes based on deviations in stress evolution.
Degree
thesis:*- Grantor dc:publisher
- University of Thessaly (UTH)
- Year dc:date
- 2025
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Stoikos, Pavlos
- Στόικος, Παύλος
Subjects
dc:subject × 26- Ηλεκτρομετανάστευση
- Αξιοπιστία Ολοκληρωμένων Κυκλωμάτων
- Εκθετικό Πίνακα
- Υπόχωρος Krylov
- Εξίσωση Korhonen
- Μοντέλο Πεπερασμένων Διαφορών
- Αγωγοί Διασυνδέσεων
- Ασφάλεια υλικού
- Χρόνος Aστοχίας
- Κακόβουλα κυκλώματα
- Electromigration
- VLSI reliability
- Matrix exponential
- Krylov subspace
- Korhonen’s Equation
- Finite Difference Modeling
- Interconnect trees
- Hardware security
- Time-to-Failure
- Hardware Trojans
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Επιστήμες Μηχανικού και Τεχνολογία
- Υπολογιστές, Υλικό (hardware) και Αρχιτεκτονική
- Electrical Engineering, Electronic Engineering, Information Engineering
- Engineering and Technology
- Computer science, Hardware and Architecture
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/60422