Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

SAMSEM -- A Generic and Scalable Approach for IC Metal Line Segmentation

Christian Gehrmann, Jonas Ricker, Simon Damm, Deruo Cheng, Julian Speith, Yiqiong Shi, Asja Fischer, Christof Paar

Abstract

In light of globalized hardware supply chains, the assurance of hardware components has gained significant interest, particularly in cryptographic applications and high-stakes scenarios. Identifying metal lines on scanning electron microscope (SEM) images of integrated circuits (ICs) is one essential step in verifying the absence of malicious circuitry in chips manufactured in untrusted environments. Due to varying manufacturing processes and technologies, such verification usually requires tuning parameters and algorithms for each target IC. Often, a machine learning model trained on images of one IC fails to accurately detect metal lines on other ICs. To address this challenge, we create SAMSEM by adapting Meta's Segment Anything Model 2 (SAM2) to the domain of IC metal line segmentation. Specifically, we develop a multi-scale segmentation approach that can handle SEM images of varying sizes, resolutions, and magnifications. Furthermore, we deploy a topology-based loss alongside pixel-based losses to focus our segmentation on electrical connectivity rather than pixel-level accuracy. Based on a hyperparameter optimization, we then fine-tune the SAM2 model to obtain a model that generalizes across different technology nodes, manufacturing materials, sample preparation methods, and SEM imaging technologies. To this end, we leverage an unprecedented dataset of SEM images obtained from 48 metal layers across 14 different ICs. When fine-tuned on seven ICs, SAMSEM achieves an error rate as low as 0.72% when evaluated on other images from the same ICs. For the remaining seven unseen ICs, it still achieves error rates as low as 5.53%. Finally, when fine-tuned on all 14 ICs, we observe an error rate of 0.62%. Hence, SAMSEM proves to be a reliable tool that significantly advances the frontier in metal line segmentation, a key challenge in post-manufacturing IC verification.

SAMSEM -- A Generic and Scalable Approach for IC Metal Line Segmentation

Abstract

In light of globalized hardware supply chains, the assurance of hardware components has gained significant interest, particularly in cryptographic applications and high-stakes scenarios. Identifying metal lines on scanning electron microscope (SEM) images of integrated circuits (ICs) is one essential step in verifying the absence of malicious circuitry in chips manufactured in untrusted environments. Due to varying manufacturing processes and technologies, such verification usually requires tuning parameters and algorithms for each target IC. Often, a machine learning model trained on images of one IC fails to accurately detect metal lines on other ICs. To address this challenge, we create SAMSEM by adapting Meta's Segment Anything Model 2 (SAM2) to the domain of IC metal line segmentation. Specifically, we develop a multi-scale segmentation approach that can handle SEM images of varying sizes, resolutions, and magnifications. Furthermore, we deploy a topology-based loss alongside pixel-based losses to focus our segmentation on electrical connectivity rather than pixel-level accuracy. Based on a hyperparameter optimization, we then fine-tune the SAM2 model to obtain a model that generalizes across different technology nodes, manufacturing materials, sample preparation methods, and SEM imaging technologies. To this end, we leverage an unprecedented dataset of SEM images obtained from 48 metal layers across 14 different ICs. When fine-tuned on seven ICs, SAMSEM achieves an error rate as low as 0.72% when evaluated on other images from the same ICs. For the remaining seven unseen ICs, it still achieves error rates as low as 5.53%. Finally, when fine-tuned on all 14 ICs, we observe an error rate of 0.62%. Hence, SAMSEM proves to be a reliable tool that significantly advances the frontier in metal line segmentation, a key challenge in post-manufacturing IC verification.
Paper Structure (41 sections, 2 equations, 9 figures, 4 tables)

This paper contains 41 sections, 2 equations, 9 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Our Contributions.
  3. Structure of this Paper.
  4. Technical Background and Related Work
  5. Metal Line Segmentation
  6. Related Work.
  7. Segment Anything Model 2
  8. Related Work.
  9. Topology-Based Loss Functions
  10. Related Work.
  11. Methodology
  12. Motivation
  13. Segmentation Pipeline
  14. Model [ boxtype=O, boxfill=blue!72.2580645161!cyan!60.7843137255!white, boxcolor=blue!72.2580645161!cyan!60.7843137255!white, boxlinewidth=0pt ]1 -- Segmenting Original-Size Images
  15. Model [ boxtype=O, boxfill=red!30.9803921569!yellow, boxcolor=red!30.9803921569!yellow, boxlinewidth=0pt ]2 -- Segmenting Images in Smaller Patches
  16. ...and 26 more sections

Figures (9)

  • Figure 1: Metal layer images from three different IC as well as the corresponding ground-truth mask and the image segmentation masks produced from different methods. Non-obvious ESD errors in the segmentation are marked with orange circles.
  • Figure 2: Model components of SAM2 and their interactions DBLP:conf/iclr/RaviGHHR0KRRGMP25.
  • Figure 3: [boxfill=red,boxcolor=red]A -- (\ref{['mseg::subfigure::large_error_input']}) to (\ref{['mseg::subfigure::large_error_segmentation']}) depict short circuits in the segmentation mask resulting from downscaling the input image to fit the shape expected by the image encoder.
  • Figure 4: [boxfill=red,boxcolor=red]B -- (\ref{['mseg::subfigure::inverse_error_input']}) to (\ref{['mseg::subfigure::inverse_error_segmentation']}) depict background being segmented as a metal line because of the lack of structures in the input image, resulting in FP or short circuit. [boxfill=red,boxcolor=red]C -- (\ref{['mseg::subfigure::flaky_error_input']}) to (\ref{['mseg::subfigure::flaky_error_segmentation']}) show white speckles around the correctly identified metal line in the segmentation mask, resulting in vast amounts of false positives.
  • Figure 5: Workflow of our multi-scale segmentation approach.
  • ...and 4 more figures