pith. sign in

arxiv: 2604.10743 · v3 · submitted 2026-04-12 · 💻 cs.AR

EMSpice 3: Full-chip Temperature-Aware Multiphysics Electromigration and IR-Drop Analysis

Haotian Lu , Sheldon X.-D. Tan This is my paper

Pith reviewed 2026-05-10 15:19 UTC · model grok-4.3

classification 💻 cs.AR
keywords electromigrationIR-drop analysispower-grid networksJoule heatingthermomigrationfull-chip simulationthermal profilesreliability
0
0 comments X

The pith

Spatial temperature profiles affect power-grid lifetime more than average temperature in full-chip EM-IR analysis.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper presents EMSpice 3, a framework that runs coupled electromigration, thermomigration, and IR-drop simulations on entire power-grid networks while including Joule heating and actual chip-wide temperature maps. It shows that the exact placement of hot spots can shift when and where voids form and whether the network hits failure thresholds, more than a single average temperature value would predict. This matters for processor reliability because power networks degrade under combined current, stress, and heat, and current design checks often rely on simplified thermal assumptions. The tool adds an immortality check, transient stress solving, resistance updates from voids, repeated IR-drop solves, and Monte Carlo runs for variation, plus a Krylov method that speeds up the math without much accuracy loss.

Core claim

EMSpice 3 performs full-chip multiphysics EM-IR analysis by jointly solving for Joule heating and practical spatial thermal profiles on extracted P/G netlists. It combines an immortality check, transient EM/TM stress evolution, void-induced resistance updates, repeated IR-drop recomputation, and optional Monte Carlo lifetime prediction. Results across benchmarks reveal that specific spatial temperature profiles impact P/G network lifetime more than average temperature alone, as higher-average-temperature maps can avoid 10% IR-drop failure if hotspots avoid critical paths in a RISC-V core, while gradients alter void locations and degraded branches; Monte Carlo further shows design-dependent T

What carries the argument

The iterative multiphysics loop that couples transient EM/TM stress evolution with void-induced resistance updates and repeated IR-drop recomputation, accelerated by an extended rational Krylov subspace solver.

If this is right

  • Higher average temperatures can still meet IR-drop limits if thermal hotspots are misaligned with high-current paths in designs like RISC-V cores.
  • Temperature gradients can relocate the site of the first critical void and change which wire branches experience the largest resistance increase.
  • Lifetime variation under 20% changes in diffusivity and critical stress differs sharply by design, reaching 15.8% coefficient of variation in RISC-V versus 0.0058% in ARM Cortex-A.
  • The extended rational Krylov method delivers 1.18x to 1.50x speedup on transient solves with sub-0.05% error in TTF and final-IR metrics.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Early-stage chip floorplanning could incorporate spatial thermal maps directly into EM checks rather than relying on uniform-temperature assumptions.
  • Similar location-specific thermal effects may apply to other interconnect degradation mechanisms such as stress migration in advanced nodes.
  • Design teams might need new thermal-aware reliability sign-off flows that iterate between power-grid layout and detailed temperature simulation.

Load-bearing premise

The underlying models for stress evolution and void growth remain accurate when scaled to full chips under the chosen immortality checks, resistance-update rules, and supplied spatial temperature maps.

What would settle it

Direct comparison of the framework's predicted time-to-failure and final IR-drop values against electrical and thermal measurements on fabricated test chips run under controlled non-uniform temperature gradients.

Figures

Figures reproduced from arXiv: 2604.10743 by Haotian Lu, Sheldon X.-D. Tan.

Figure 1
Figure 1. Figure 1: Stress distribution with void thickness of the copper-void boundary, δ, as infinitesimally small. We then establish the stress gradient between the zero￾stress void surface and the adjacent metal as indicated in [3], [19]: ∂σ ∂x [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: summarizes the overall analysis flow as implemented in EMSpice 3. The analysis starts from three inputs: an extracted power-grid netlist, a YAML parameter file containing the material and numerical settings, and an optional thermal map for TM-aware analysis. After parsing these files, the framework can perform an optional steady-state screening pass to prune trees that are already classified as immortal. P… view at source ↗
Figure 3
Figure 3. Figure 3: Mid-tree void nucleation in the Dual-Port RAM design. Left: stress profile along tree (25-100) over outer timesteps from 2.5×106 s to 5.0 × 107 s, showing the evolution from a near-zero state to a strong compressive gradient as degradation accumulates. Right: stress history at junction node 220, where the abrupt transition from about 3.3×108 Pa tensile stress to about −3.4×108 Pa compressive stress marks v… view at source ↗
Figure 5
Figure 5. Figure 5: shows the steady-state stress distribution for Case 1 computed without void nucleation. Without the saturation mechanism, stress accumulates continuously and peak tensile values in the lower-left hotspot far exceed those in Fig. 8a, where void nucleation clamps stress near the saturation thresh￾old once flux is interrupted. This comparison highlights the importance of accurately modeling nucleation in coup… view at source ↗
Figure 4
Figure 4. Figure 4: RISC-V core: finished physical layout (left) and extracted power grid structure showing four metal layers (M1, M2, M7, M8) with VIA connections (right). We run two scenarios that differ only in their spatial temperature distribution, both with the same 353 K average temperature. Case 1 — Baseline external temperature map. The first scenario uses a baseline spatial temperature distribution provided external… view at source ↗
Figure 6
Figure 6. Figure 6: Externally given spatial temperature maps provided as inputs to EMSpice 3 for the two RISC-V core scenarios. Both maps share a 353 K die-average temperature but differ in spatial structure: the baseline map has a compact lower-left hotspot; the Qualcomm measured map spreads heat broadly across the central die area. (a) Per-node Joule-heating tem￾perature distribution (Case 1): in￾tense heating along high-c… view at source ↗
Figure 7
Figure 7. Figure 7: Per-node Joule-heating temperature distributions on the RISC-V power grid computed by EMSpice 3 for the two thermal scenarios. Individual wire segments are resolved, showing the spatial correlation between current density and local temperature. In Case 1 the external baseline hotspot directly overlaps the most current￾stressed region; in Case 2 the Qualcomm ambient heat does not preferentially load the hig… view at source ↗
Figure 8
Figure 8. Figure 8: Stress distributions and IR-drop maps for the RISC-V core under the two thermal scenarios (external input maps [PITH_FULL_IMAGE:figures/full_fig_p010_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Power grid structure of the ARM Cortex-A logic core (three metal layers M1, M8, M9 shown together with VIA connections) the design. The initial maximum IR-drop is 8.85% of Vsrc; it rises to 22.21% by the end of the simulation. The 10% IR￾drop failure threshold is crossed at approximately 1.05×107 s (≈4.0 months), as shown in Fig. 10e, where the maximum￾drop node lies near the left boundary of the grid. Cas… view at source ↗
Figure 10
Figure 10. Figure 10: Temperature maps, stress distributions, and IR-drop maps for the ARM Cortex-A logic core under two thermal scenarios we perform Monte Carlo (MC) simulations on both the RISC￾V core and the ARM Cortex-A logic core designs. In each MC run, the EM diffusivity κ(x) and critical stress parameters are randomly perturbed with a coefficient of variation (CoV) of 20%. For each design, 100 independent MC samples ar… view at source ↗
Figure 11
Figure 11. Figure 11: 3D wire-level temperature distribution of the ARMcore power grid under uniform 353.0 K ambient temperature with Joule heating only and no external thermal-map input. The global field stays near the ambient baseline, while localized hot segments produce sharp thermal peaks over the interconnect network. (a) Tree 11-133 (b) Tree 18-122 (c) Tree 18-128 (d) Tree 19-101 [PITH_FULL_IMAGE:figures/full_fig_p013_… view at source ↗
Figure 12
Figure 12. Figure 12: Representative node-index temperature profiles for four ARMcore trees under uniform 353.0 K ambient temperature with Joule heating only and no external thermal-map input. The profiles highlight localized intra-tree hot spots superimposed on a near￾uniform 353 K baseline. standard deviation of 0.1258 years (≈ 3.969×106 s), yielding CoV of 15.77%. These statistics are computed only from the uncensored failu… view at source ↗
Figure 14
Figure 14. Figure 14: Monte Carlo TTF distribution for the ARM Cortex-A logic core (100 runs; 20% variation in κ(x) and critical stress). Mean TTF = 0.3307 years, σ = 1.92 × 10−5 years; 0 censored runs. spatial temperature structure and the margin of individual trees relative to the nucleation threshold must be characterized for predictive full-chip EM reliability analysis. VI. CONCLUSIONS We presented EMSpice 3, a full-chip m… view at source ↗
read the original abstract

This paper presents EMSpice~3, a full-chip multiphysics framework for coupled electromigration (EM), thermomigration (TM), and IR-drop analysis of practical power-grid (P/G) networks. The framework is, to our knowledge, the first EM-IR analysis flow that jointly incorporates Joule heating and practical spatial thermal profiles for full-chip P/G network designs. It operates on extracted power-grid netlists and combines an immortality check, transient EM/TM stress evolution, void-induced resistance updates, repeated IR-drop recomputation, and optional Monte Carlo lifetime prediction. To make chip-level EM analysis tractable, the framework integrates an extended rational Krylov subspace method into the transient solver, achieving $1.18\times$--$1.50\times$ speedup with sub-0.05% reported TTF/final-IR metric error relative to the default non-Krylov FDTD analysis across six benchmark designs. The numerical results reveal that the specific spatial temperature profile can have a more significant impact on P/G network lifetime than the average temperature itself. In the RISC-V core, a higher-average-temperature profile can avoid the 10% IR-drop failure threshold when its hotspots are less aligned with critical current paths, while mapped temperature gradients can move the critical void location and change which resistor branches are degraded. Monte Carlo analysis further shows design-specific variation sensitivity: under 20% variation in EM diffusivity and critical stress, the RISC-V core exhibits about 15.8% TTF coefficient of variation, whereas the ARM Cortex-A logic core exhibits only 0.0058\%. These results show that practical thermal profiles, resistance feedback, and stochastic material variation must be considered jointly for predictive full-chip EM-IR analysis.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript presents EMSpice 3, a full-chip multiphysics framework for coupled electromigration (EM), thermomigration (TM), and IR-drop analysis of power-grid networks. It integrates Joule heating with spatial thermal profiles, an immortality check, transient stress evolution, void-induced resistance updates, repeated IR-drop recomputation, and an extended rational Krylov subspace method for the transient solver. On six benchmark designs (including RISC-V and ARM cores), it reports 1.18–1.50× speedup with sub-0.05% TTF and final-IR error versus non-Krylov FDTD, shows that specific spatial temperature profiles can affect lifetime more than average temperature, and provides Monte Carlo results on TTF variation under material parameter uncertainty.

Significance. If the underlying EM/TM models and supplied thermal profiles remain predictive, the work would be significant for enabling practical full-chip temperature-aware reliability analysis. The reported efficiency gains from the Krylov integration and the demonstration of spatial thermal effects versus average temperature would highlight the need to incorporate realistic thermal maps and stochastic variation in P/G design flows.

major comments (3)
  1. [Abstract] Abstract: The 1.18–1.50× speedup and sub-0.05% TTF/IR metric error are presented as evidence of tractability, but no derivation, bound, or explicit error analysis for the extended rational Krylov subspace approximation is provided; this is load-bearing for the central efficiency claim.
  2. [Numerical results] Numerical results: The headline conclusion that specific spatial temperature profiles impact P/G network lifetime more than average temperature (e.g., RISC-V core avoiding 10% IR-drop threshold via hotspot alignment) rests on the fidelity of the EM/TM stress-evolution equations, immortality check, and void-induced resistance updates when coupled to IR-drop recomputation; the manuscript contains no silicon correlation data, calibration against measured thermal maps, or validation of these components at full-chip scale.
  3. [Monte Carlo analysis] Monte Carlo analysis: The reported TTF coefficients of variation (15.8% for RISC-V, 0.0058% for ARM Cortex-A) under 20% variation in EM diffusivity and critical stress treat these percentages as given free parameters without a sensitivity study or justification tied to process data, which directly affects the stochastic lifetime prediction results.
minor comments (2)
  1. [Abstract] Abstract: The notation 'EMSpice~3' uses a non-standard tilde; adopt consistent formatting throughout.
  2. [Abstract] Abstract: The 'to our knowledge' claim of being the first EM-IR flow to jointly incorporate Joule heating and practical spatial thermal profiles would be strengthened by a short literature comparison table in the introduction.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed comments. We address each major point below, indicating planned revisions to strengthen the manuscript.

read point-by-point responses

  1. Referee: [Abstract] The 1.18–1.50× speedup and sub-0.05% TTF/IR metric error are presented as evidence of tractability, but no derivation, bound, or explicit error analysis for the extended rational Krylov subspace approximation is provided; this is load-bearing for the central efficiency claim.

    Authors: We agree that explicit error analysis is needed to support the efficiency claims. The extended rational Krylov method extends established model-order reduction techniques with known convergence properties. In the revision we will add a dedicated subsection deriving the approximation error bounds from rational Krylov subspace theory, with supporting references, to justify the reported sub-0.05% metric error. revision: yes

  2. Referee: [Numerical results] The headline conclusion that specific spatial temperature profiles impact P/G network lifetime more than average temperature rests on the fidelity of the EM/TM stress-evolution equations, immortality check, and void-induced resistance updates when coupled to IR-drop recomputation; the manuscript contains no silicon correlation data, calibration against measured thermal maps, or validation of these components at full-chip scale.

    Authors: This observation is correct. The framework employs physics-based EM/TM models from the established literature and realistic thermal profiles from chip floorplans, but the manuscript provides no silicon correlation or measured-map calibration. We will add an explicit limitations paragraph in the discussion section stating the modeling assumptions and noting that full-chip experimental validation lies beyond the present scope. Relative comparisons between thermal profiles remain internally consistent under the same equations. revision: partial

  3. Referee: [Monte Carlo analysis] The reported TTF coefficients of variation (15.8% for RISC-V, 0.0058% for ARM Cortex-A) under 20% variation in EM diffusivity and critical stress treat these percentages as given free parameters without a sensitivity study or justification tied to process data, which directly affects the stochastic lifetime prediction results.

    Authors: The 20% variation was selected as representative of typical process-induced spreads in EM parameters. We will incorporate a sensitivity study varying the percentage from 10% to 30% and add citations to semiconductor process-variation literature to justify the range, thereby demonstrating robustness of the design-specific TTF variation results. revision: yes

Circularity Check

0 steps flagged

No significant circularity in framework application to external benchmarks

full rationale

The paper presents a simulation framework that applies established EM/TM stress-evolution models, immortality checks, void-resistance updates, and an extended rational Krylov solver to extracted netlists and supplied spatial thermal profiles from independent benchmark designs (RISC-V, ARM cores). Reported TTF, IR-drop, and speedup metrics are direct simulation outputs on these external inputs rather than quantities fitted or defined inside the paper and then re-predicted. No equations reduce the central claims about spatial temperature impact or Monte Carlo variation to self-referential constructions, and self-citations to prior EMSpice solver components describe reusable numerical methods without making the comparative lifetime results circular by construction. The derivation chain is therefore self-contained against the provided benchmarks.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 0 invented entities

The framework rests on standard electromigration physics models whose accuracy at full-chip scale is assumed rather than re-derived; the only explicit free parameters appear in the Monte Carlo variation study.

free parameters (2)
  • EM diffusivity variation percentage
    20 percent variation assumed in Monte Carlo runs to compute TTF coefficient of variation on the RISC-V and ARM cores.
  • critical stress variation percentage
    20 percent variation assumed alongside diffusivity in the same Monte Carlo lifetime prediction.
axioms (2)
  • domain assumption Standard continuum models for electromigration stress evolution and void nucleation remain valid when applied to full-chip extracted netlists with spatially varying temperature.
    Invoked throughout the transient solver and immortality check without additional justification in the abstract.
  • domain assumption The extended rational Krylov subspace method produces sub-0.05 percent error in TTF and final IR-drop metrics relative to full FDTD.
    Used to justify the speedup claim; error bound is reported but not derived in the provided abstract.

pith-pipeline@v0.9.0 · 5628 in / 1669 out tokens · 49293 ms · 2026-05-10T15:19:20.791848+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

25 extracted references · 25 canonical work pages

  1. [1]

    Electromigration-A Brief Survey and Some Recent Re- sults,

    J. R. Black, “Electromigration-A Brief Survey and Some Recent Re- sults,”IEEE Trans. on Electron Devices, vol. 16, no. 4, pp. 338–347, Apr. 1969

    work page 1969

  2. [2]

    Electromigration in thin aluminum films on titanium nitride,

    I. A. Blech, “Electromigration in thin aluminum films on titanium nitride,”Journal of Applied Physics, vol. 47, no. 4, pp. 1203–1208, 1976

    work page 1976

  3. [3]

    Stress evolution due to electromigration in confined metal lines,

    M. A. Korhonen, P. Borgesen, K. N. Tu, and C. Y . Li, “Stress evolution due to electromigration in confined metal lines,”Journal of Applied Physics, vol. 73, no. 8, pp. 3790–3799, 1993

    work page 1993

  4. [4]

    Fast electromigration stress evolution analysis for interconnect trees using krylov subspace method,

    C. Cook, Z. Sun, E. Demircan, M. D. Shroff, and S. X.-D. Tan, “Fast electromigration stress evolution analysis for interconnect trees using krylov subspace method,”IEEE Trans. on Very Large Scale Integration (VLSI) Systems, vol. 26, no. 5, pp. 969–980, May 2018

    work page 2018

  5. [5]

    An- alytical modeling and characterization of electromigration effects for multibranch interconnect trees,

    H. Chen, S. X.-D. Tan, X. Huang, T. Kim, and V . Sukharev, “An- alytical modeling and characterization of electromigration effects for multibranch interconnect trees,”IEEE Trans. on Computer-Aided Design of Integrated Circuits and Systems, vol. 35, no. 11, pp. 1811–1824, 2016

    work page 2016

  6. [6]

    RAIN: A tool for reliability assessment of interconnect networks—physics to software,

    A. Abbasinasab and M. Marek-Sadowska, “RAIN: A tool for reliability assessment of interconnect networks—physics to software,” inProc. Design Automation Conf. (DAC). New York, NY , USA: ACM, 2018, pp. 133:1–133:6

    work page 2018

  7. [7]

    A fast semi-analytic approach for combined electromigration and ther- momigration analysis for general multisegment interconnects,

    L. Chen, S. X.-D. Tan, Z. Sun, S. Peng, M. Tang, and J. Mao, “A fast semi-analytic approach for combined electromigration and ther- momigration analysis for general multisegment interconnects,”IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, vol. 40, no. 2, pp. 350–363, 2021

    work page 2021

  8. [8]

    A study of tapered 3-d TSVs for power and thermal integrity,

    A. Todri, S. Kundu, P. Girard, A. Bosio, L. Dilillo, and A. Virazel, “A study of tapered 3-d TSVs for power and thermal integrity,”IEEE Transactions on Very Large Scale Integration (VLSI) Systems, vol. 21, no. 2, pp. 306–319, Feb 2013

    work page 2013

  9. [9]

    Fast electromigration stress analysis considering spatial joule heating effects,

    M. Kavousi, L. Chen, and S. X.-D. Tan, “Fast electromigration stress analysis considering spatial joule heating effects,” in2022 27th Asia and South Pacific Design Automation Conference (ASP-DAC), 2022, pp. 208–213

    work page 2022

  10. [10]

    EMSpice: Physics- Based Electromigration Check Using Coupled Electronic and Stress Simulation,

    Z. Sun, S. Yu, H. Zhou, Y . Liu, and S. X.-D. Tan, “EMSpice: Physics- Based Electromigration Check Using Coupled Electronic and Stress Simulation,”IEEE Transactions on Device and Materials Reliability, vol. 20, no. 2, pp. 376–389, Jun. 2020

    work page 2020

  11. [11]

    EMspice 2.0: Multi- physics electromigration analysis tool for beyond moore ics,

    S. Lamichhane, M. Kavousi, and S. X.-D. Tan, “EMspice 2.0: Multi- physics electromigration analysis tool for beyond moore ics,” in2024 IEEE Computer Society Annual Symposium on VLSI (ISVLSI), 2024

    work page 2024

  12. [12]

    Proton – a python framework for physics-based electromigration as- sessment on contemporary vlsi power grids,

    O. Axelou, E. Tselepi, G. Floros, N. Evmorfopoulos, and G. Stamoulis, “Proton – a python framework for physics-based electromigration as- sessment on contemporary vlsi power grids,” in2023 19th International Conference on Synthesis, Modeling, Analysis and Simulation Methods and Applications to Circuit Design (SMACD), 2023, pp. 1–4

    work page 2023

  13. [13]

    A fast semi-analytic approach for combined electromigration and ther- momigration analysis for general multi-segment interconnects,

    L. Chen, S. X.-D. Tan, Z. Sun, S. Peng, M. Tang, and J. Mao, “A fast semi-analytic approach for combined electromigration and ther- momigration analysis for general multi-segment interconnects,”IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, pp. 1–1, 2020

    work page 2020

  14. [14]

    H. S. Carslaw and J. C. Jaeger,Conduction of Heat in Solids, 2nd ed. Oxford: Oxford University Press, 1959

    work page 1959

  15. [15]

    Electromigration Immortality Check considering Joule Heating Effect for Multisegment Wires,

    M. Kavousi, L. Chen, and S. X.-D. Tan, “Electromigration Immortality Check considering Joule Heating Effect for Multisegment Wires,” in Proc. Int. Conf. on Computer Aided Design (ICCAD), 2020, pp. 1–8

    work page 2020

  16. [16]

    Recent advances in EM and BTI induced reliability modeling, analysis and optimization,

    S. X.-D. Tan, H. Amrouch, T. Kim, Z. Sun, C. Cook, and J. Henkel, “Recent advances in EM and BTI induced reliability modeling, analysis and optimization,”Integration, the VLSI Journal, vol. 60, pp. 132–152, Jan. 2018

    work page 2018

  17. [17]

    S. X.-D. Tan, M. Tahoori, T. Kim, S. Wang, Z. Sun, and S. Kiamehr, VLSI Systems Long-Term Reliability – Modeling, Simulation and Opti- mization. Springer Publishing, 2019

    work page 2019

  18. [18]

    Physically based models of electromigration: From black’s equation to modern tcad models,

    R. de Orio, H. Ceric, and S. Selberherr, “Physically based models of electromigration: From black’s equation to modern tcad models,” Microelectronics Reliability, vol. 50, no. 6, pp. 775–789, 2010, 2009 Reliability of Compound Semiconductors (ROCS) Workshop). [Online]. Available: https://www.sciencedirect.com/science/article/pii/ S0026271410000193

    work page 2010

  19. [19]

    Postvoiding stress evolution in confined metal lines,

    V . Sukharev, A. Kteyan, and X. Huang, “Postvoiding stress evolution in confined metal lines,”IEEE Transactions on Device and Materials Reliability, vol. 16, no. 1, pp. 50–60, 2016

    work page 2016

  20. [20]

    Reliability computer-aided design tool for full-chip electromigration analysis and comparison with different interconnect metallizations,

    S. M. Alam, C. L. Gan, C. V . Thompson, and D. E. Troxel, “Reliability computer-aided design tool for full-chip electromigration analysis and comparison with different interconnect metallizations,”Microelectron. J., vol. 38, no. 4-5, pp. 463–473, Apr. 2007. [Online]. Available: http://dx.doi.org/10.1016/j.mejo.2006.11.017

    work page doi:10.1016/j.mejo.2006.11.017 2007

  21. [21]

    Novel physics-based tool-prototype for electromigration assessment in commercial-grade power delivery networks,

    S. Torosyan, A. Kteyan, V . Sukharev, J.-H. Choy, and F. N. Najm, “Novel physics-based tool-prototype for electromigration assessment in commercial-grade power delivery networks,”Journal of Vacuum Science & Technology B, vol. 39, no. 1, p. 013203, 12 2020. [Online]. Available: https://doi.org/10.1116/6.0000617

    work page doi:10.1116/6.0000617 2020

  22. [22]

    Accelerating physics-based electromigration analysis via rational krylov subspaces,

    H. Lu and S. X.-D. Tan, “Accelerating physics-based electromigration analysis via rational krylov subspaces,”arXiv preprint arXiv:2602.00330, Feb 2026. [Online]. Available: https://arxiv.org/abs/2602.00330

    work page arXiv 2026

  23. [23]

    Fast Physics-Based Electromigration Checking for On-Die Power Grids,

    S. Chatterjee, V . Sukharev, and F. N. Najm, “Fast Physics-Based Electromigration Checking for On-Die Power Grids,” inProceedings of the 35th International Conference on Computer-Aided Design, ser. ICCAD ’16. New York, NY: ACM Press, Nov. 2016, pp. 1–8. [Online]. Available: http://doi.acm.org/10.1145/2966986.2967041

    work page doi:10.1145/2966986.2967041 2016

  24. [24]

    Synopsys 32/28nm generic library for teaching ic design,

    Synopsys, “Synopsys 32/28nm generic library for teaching ic design,” Accessed: Feb. 2019, [Online]. Available. [Online]. Available: http://www.synopsys.com

    work page 2019

  25. [25]

    Commerical Multi/Many Cores Thermal Map Dataset,

    “Commerical Multi/Many Cores Thermal Map Dataset,” https://github. com/sheldonucr/commercial\ thermal\ map\ dataset

    work page