Pith. sign in

REVIEW 1 major objections 1 minor 41 references

2.5D Root of Trust: Securing the Chiplet Ecosystem

T0 review · 1 major / 1 minor · reviewed 2026-06-26 · grok-4.3

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read An active interposer can act as a physically isolated 2.5D root of trust that enforces memory permissions and blocks coherence attacks on commodity chiplets without any changes to those chiplets.

desk verdict The paper sketches a conceptual 2.5D Root of Trust via an active interposer but leaves the trust model and enforcement mechanisms at a high level without supporting analysis. read the letter →

arxiv 2606.22198 v1 pith:QO6YXM5R submitted 2026-06-20 cs.CR cs.AR

Charles Williams , Mohammed Nabeel , Gino Chacon , Ozgur Sinanoglu , Paul V. Gratz , Johann Knechtel This is my paper
classification cs.CRcs.AR
keywords chipletsecurity2.5DintegrationrootoftrusthardwareTrojanscoherenceattacksinterposerinterconnectsplitmanufacturing
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper examines how 2.5D chiplet systems assembled via silicon interposers create new attack surfaces when chiplets come from untrusted sources. It surveys threats including interconnect snooping and spoofing, cache coherence forging attacks, and microarchitectural side channels. The central proposal positions the interposer itself as the trusted component by embedding transaction monitors and coherence message checkers directly in its fabric. These monitors enforce access permissions at the hardware level during runtime and neutralize coherence exploits by construction. The approach also reviews associated EDA flows that can simultaneously improve power, signal integrity, and system footprint.

What carries the argument

The active interposer fabric configured as a physically isolated 2.5D root of trust, with embedded transaction monitors and coherence message checkers that inspect and gate all inter-chiplet traffic.

What would settle it

Demonstration of a coherence forging attack or unauthorized memory access that succeeds despite the presence of interposer-embedded monitors and checkers.

Watch

Extended reading notes

Core claim

By embedding transaction monitors and coherence message checkers within the trusted interposer fabric, the system enforces memory access permissions by construction and neutralizes coherence-level attacks without need for modifying or securing the commodity chiplets.

Load-bearing premise

The interposer can be fabricated and operated as a fully trusted component even when sourced from potentially untrusted vendors.

Editorial extensions

If this is right

  • Memory access permissions become enforced directly by the interposer hardware rather than by the chiplets.
  • Coherence-level attacks are blocked at the message level without requiring changes inside individual chiplets.
  • The same interposer defenses address interconnect snooping, spoofing, and man-in-the-middle attacks.
  • EDA flows for the interposer can be used to reduce overall system power and improve signal integrity.
  • Heterogeneous multi-vendor chiplet systems become viable without securing every participating die.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same monitor placement could extend to other 2.5D or 3D integration schemes where a central fabric is already present.
  • Designers could prioritize interposer trust over per-chiplet security features when sourcing parts from multiple foundries.
  • Runtime monitoring in the interposer might also limit the effectiveness of certain microarchitectural side channels that rely on shared coherence traffic.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

1 major / 1 minor

Summary. The manuscript is a perspective paper surveying security threats in 2.5D chiplet ecosystems (interconnect attacks, cache coherence exploits, and microarchitectural side channels) and proposing an active interposer as a physically isolated 2.5D Root of Trust. The central claim is that embedding transaction monitors and coherence message checkers in the interposer fabric enforces memory access permissions by construction and neutralizes coherence-level attacks without modifying commodity chiplets; design-time 2.5D split manufacturing is examined to address untrusted vendors for any component including the interconnect, and EDA flows are reviewed that also improve power, signal integrity, and footprint.

Significance. If the architectural proposal holds, it offers a concrete way to concentrate trust in the interposer for heterogeneous multi-vendor systems, reducing the attack surface on individual chiplets. The survey of threat categories and existing strategies provides a useful framing for the field. The paper explicitly credits split manufacturing as a design-time complement to runtime isolation and notes concurrent benefits to power and signal integrity.

major comments (1)
  1. [Abstract] Abstract: the central claim that the active interposer can be positioned as a 'physically isolated 2.5D RoT' that 'enforces memory access permissions by construction' rests on the assumption that split manufacturing suffices to prevent subversion of the embedded transaction monitors and coherence checkers even when the interconnect fabric originates from an untrusted vendor. No concrete mechanism, threat model, or verification argument is supplied showing that the monitors themselves remain sound under this supply-chain constraint.
minor comments (1)
  1. The phrase 'so-called transaction monitors' is introduced without a definition or forward reference; a one-sentence description of their intended function would aid readability.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for this constructive comment on the abstract. The observation correctly identifies that our high-level positioning of the active interposer as a 2.5D RoT rests on an implicit assumption about split manufacturing that is not elaborated with mechanisms or arguments in the current text. We will revise accordingly.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that the active interposer can be positioned as a 'physically isolated 2.5D RoT' that 'enforces memory access permissions by construction' rests on the assumption that split manufacturing suffices to prevent subversion of the embedded transaction monitors and coherence checkers even when the interconnect fabric originates from an untrusted vendor. No concrete mechanism, threat model, or verification argument is supplied showing that the monitors themselves remain sound under this supply-chain constraint.

    Authors: We agree that the manuscript, as a perspective paper, presents the active interposer proposal at an architectural level and does not supply a concrete mechanism, threat model, or verification argument demonstrating that split manufacturing alone suffices to protect the embedded monitors and checkers. The text treats split manufacturing as a complementary design-time defense that, together with the runtime isolation properties of the interposer, enables the RoT claim, but this dependency is not made explicit. We will revise the abstract to qualify the claim (e.g., “under the assumption of trusted interposer fabrication via split manufacturing”) and will add a short discussion paragraph noting that detailed supply-chain verification of the interposer remains an open research question. This change will align the stated claims with the scope and depth of the paper. revision: yes

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: conceptual proposal without derivations or self-referential steps

full rationale

The paper is a perspective survey on 2.5D chiplet security threats and defenses. It presents no equations, fitted parameters, predictions, or derivation chains. The central proposal—that an active interposer can serve as a trusted RoT by embedding monitors and checkers—rests on stated architectural assumptions about physical isolation and split manufacturing rather than any reduction to prior inputs by construction. No self-citations are invoked as load-bearing uniqueness theorems, and no ansatzes or renamings of known results appear. The absence of mathematical or empirical modeling means the circularity patterns do not apply.

Assumptions & free parameters 0 free parameters · 1 assumptions · 1 invented entities

The proposal rests on the domain assumption that a physically isolated interposer can be trusted and that embedding monitors does not create new attack surfaces; no free parameters or invented entities with independent evidence are introduced.

assumptions (1)
  • domain assumption The interposer can be manufactured and operated as a trusted component separate from untrusted chiplets.
    Invoked when positioning the active interposer as the physically isolated 2.5D RoT that enforces permissions by construction.
invented entities (1)
  • 2.5D Root of Trust via active interposer
    purpose: To serve as the trusted enforcement point for memory permissions and coherence checks.
    Proposed as the core defense mechanism; no independent evidence or falsifiable prediction outside the paper is supplied.

how reviews work

0 comments
Cite this review

Pith. "Pith review of 2.5D Root of Trust: Securing the Chiplet Ecosystem." pith.science (2026). https://pith.science/paper/QO6YXM5R

@misc{pith2026260622198,
  author       = {Pith},
  title        = {Pith review of: 2.5D Root of Trust: Securing the Chiplet Ecosystem},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QO6YXM5R}},
  note         = {Machine review of arXiv:2606.22198}
}
read the original abstract

The semiconductor industry is rapidly transitioning from monolithic systems-on-chip toward heterogeneous, multi-vendor 2.5D chiplet ecosystems integrated via silicon interposers. While this paradigm shift offers immense benefits in yield, cost, and time-to-market, it radically expands the attack surface. Integrating chiplets from untrusted foundries and design houses introduces vulnerabilities to hardware Trojans, IP piracy, and system-level communication exploits. Critically, chip-level security features and conventional Root of Trust (RoT) proposals are insufficient in this context: any component, including the interconnect fabric itself, may be sourced from an untrusted vendor. This perspective paper surveys state-of-the-art security strategies for interposer-based 2.5D integration, focusing on three threat categories: interconnect attacks (snooping, spoofing, and man-in-the-middle), cache coherence exploits including complex forging attacks, and microarchitectural side-channel threats. We examine design-time defenses via 2.5D split manufacturing and, more crucially, runtime defenses that establish an active interposer as a physically isolated 2.5D RoT. By embedding so-called transaction monitors and coherence message checkers within the trusted interposer fabric, the system enforces memory access permissions by construction and neutralizes coherence-level attacks without need for modifying/securing the commodity chiplets. Finally, we review the EDA flows required to realize these defenses and show they concurrently improve power and signal integrity while reducing overall system footprint.

Figures

Figures reproduced from arXiv: 2606.22198 by the authors.

Figure 1
Figure 1. Overview of key concepts and findings of our prior work reviewed [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

41 extracted references · 8 canonical work pages

  1. [1]

    Understanding chiplets today to anticipate future integration opportunities and limits,

    G. H. Loh, S. Naffziger, and K. Lepak, “Understanding chiplets today to anticipate future integration opportunities and limits,” inProc. DATE. IEEE, 2021, pp. 142–145

    2021

  2. [2]

    Architecture, chip, and package co- design flow for 2.5 d ic design enabling heterogeneous ip reuse,

    J. Kim, G. Murali, H. Park, E. Qin, H. Kwon, V . Chaitanya, K. Chekuri, N. Dasari, A. Singh, M. Leeet al., “Architecture, chip, and package co- design flow for 2.5 d ic design enabling heterogeneous ip reuse,” in Proc. DAC, 2019, pp. 1–6

    2019

  3. [3]

    2.2 amd chiplet architecture for high-performance server and desktop products,

    S. Naffziger, K. Lepak, M. Paraschou, and M. Subramony, “2.2 amd chiplet architecture for high-performance server and desktop products,” inProc. ISSCC, 2020, pp. 44–45

    2020

  4. [4]

    Pioneering chiplet technology and design for the amd epyc™ and ryzen™ processor families: Industrial product,

    S. Naffziger, N. Beck, T. Burd, K. Lepak, G. H. Loh, M. Subramony, and S. White, “Pioneering chiplet technology and design for the amd epyc™ and ryzen™ processor families: Industrial product,” inProc. ISCA. IEEE, 2021, pp. 57–70

    2021

  5. [5]

    Apple unveils M1 Ultra, the world’s most powerful chip for a personal computer,

    Apple Inc., “Apple unveils M1 Ultra, the world’s most powerful chip for a personal computer,” Apple Newsroom, Mar. 2022. [Online]. Available: www.apple.com/newsroom/2022/03/ apple-unveils-m1-ultra-the-worlds-most-powerful-chip-for-a-personal-computer/

    2022

  6. [6]

    Amd instinct™ mi300x: a generative ai accelerator and platform architecture,

    A. Smith and V . K. Alla, “Amd instinct™ mi300x: a generative ai accelerator and platform architecture,”IEEE Micro, 2025

    2025

  7. [7]

    Sapphire rapids: The next-generation intel xeon scalable processor,

    N. Nassif, A. O. Munch, C. L. Molnar, G. Pasdast, S. V . Lyer, Z. Yang, O. Mendoza, M. Huddart, S. Venkataraman, S. Kandulaet al., “Sapphire rapids: The next-generation intel xeon scalable processor,” inProc. ISSCC, vol. 65, 2022, pp. 44–46

    2022

  8. [8]

    XeHPC Ponte Vecchio ,

    D. Blythe, “ XeHPC Ponte Vecchio ,” inProc. HotChips, 2021, pp. 1–34. [Online]. Available: https://doi.ieeecomputersociety.org/10.1109/ HCS52781.2021.9567038

    work page arXiv 2021

Show all 41 references
  1. [9]

    3.3 a 14nm 1ghz fpga with 2.5 d transceiver integration,

    D. Greenhill, R. Ho, D. Lewis, H. Schmit, K. H. Chan, A. Tong, S. Atsatt, D. How, P. McElheny, K. Duwelet al., “3.3 a 14nm 1ghz fpga with 2.5 d transceiver integration,” inProc. ISSCC, 2017, pp. 54– 55

    2017

  2. [10]

    2.5d root of trust: Secure system-level integration of untrusted chiplets,

    M. Nabeel, M. Ashraf, S. Patnaik, V . Soteriou, O. Sinanoglu, and J. Knechtel, “2.5d root of trust: Secure system-level integration of untrusted chiplets,”IEEE Transactions on Computers, vol. 69, no. 11, pp. 1611–1625, 2020

    2020

  3. [11]

    Coherence attacks and countermeasures in interposer- based chiplet systems,

    G. A. Chacon, C. Williams, J. Knechtel, O. Sinanoglu, P. V . Gratz, and V . Soteriou, “Coherence attacks and countermeasures in interposer- based chiplet systems,”ACM Transactions on Architecture and Code Optimization (TACO), vol. 21, no. 2, pp. 1–25, 2024

    2024

  4. [12]

    Security technology building a secure system using trustzone technology (white paper),

    A. ARM, “Security technology building a secure system using trustzone technology (white paper),”ARM Limited, 2009

    2009

  5. [13]

    Intel® software guard extensions (intel® sgx) support for dynamic memory management inside an enclave,

    F. McKeen, I. Alexandrovich, I. Anati, D. Caspi, S. Johnson, R. Leslie- Hurd, and C. Rozas, “Intel® software guard extensions (intel® sgx) support for dynamic memory management inside an enclave,” inProc. Hardware and Architectural Support for Security and Privacy, 2016, pp. 1–9

    2016

  6. [14]

    Securing system-on-chips: Hardware protection in the age of chiplets,

    L. Liu, “Securing system-on-chips: Hardware protection in the age of chiplets,” PUFsecurity, May 2023, accessed: Apr. 11,

    2023

  7. [15]

    Available: https://www.pufsecurity.com/document/ securing-system-on-chips-hardware-protection-in-the-age-of-chiplets/

    [Online]. Available: https://www.pufsecurity.com/document/ securing-system-on-chips-hardware-protection-in-the-age-of-chiplets/

  8. [16]

    From monolithic SoCs to chiplets: A new hardware security paradigm,

    Rambus, “From monolithic SoCs to chiplets: A new hardware security paradigm,” Semiconductor Engineering, Feb. 2026, accessed: Apr. 11, 2026. [Online]. Available: https://semiengineering.com/ from-monolithic-socs-to-chiplets-a-new-hardware-security-paradigm/

    2026

  9. [17]

    Packet header attack by hardware trojan in noc based tcmp and its impact analysis,

    V . J. Kulkarni, R. Manju, R. Gupta, J. Jose, and S. Nandi, “Packet header attack by hardware trojan in noc based tcmp and its impact analysis,” inProc. 15th Int. Symp. on Networks-on-Chip, ser. NOCS ’21, 2021, p. 21–28. [Online]. Available: https://doi.org/10.1145/3479876.3481597

    2021 doi

  10. [18]

    A survey of network-on-chip security attacks and countermeasures,

    S. Charles and P. Mishra, “A survey of network-on-chip security attacks and countermeasures,”ACM Comput. Surv., vol. 54, no. 5, May 2021. [Online]. Available: https://doi.org/10.1145/3450964

    2021 doi

  11. [19]

    3d integration: Another dimension toward hardware security,

    J. Knechtel, S. Patnaik, and O. Sinanoglu, “3d integration: Another dimension toward hardware security,” inProc. IOLTS, 2019, pp. 147– 150

    2019

  12. [20]

    Large-scale 3d chips: Challenges and solutions for design automation, testing, and trustworthy integration,

    J. Knechtel, O. Sinanoglu, I. A. M. Elfadel, J. Lienig, and C. C. N. Sze, “Large-scale 3d chips: Challenges and solutions for design automation, testing, and trustworthy integration,”IPSJ Transactions on System and LSI Design Methodology, vol. 10, pp. 45–62, 2017

    2017

  13. [21]

    On mitigation of side-channel attacks in 3d ics: Decorrelating thermal patterns from power and activity,

    J. Knechtel and O. Sinanoglu, “On mitigation of side-channel attacks in 3d ics: Decorrelating thermal patterns from power and activity,” inProc. DAC, 2017. [Online]. Available: https: //doi.org/10.1145/3061639.3062293

    2017 doi

  14. [22]

    The hardware trojan war,

    S. Bhunia and M. Tehranipoor, “The hardware trojan war,”Cham,, Switzerland: Springer, 2018

    2018

  15. [23]

    Hardware trojan threats to cache coherence in modern 2.5d chiplet systems,

    G. A. Chacon, C. Williams, J. Knechtel, O. Sinanoglu, and P. V . Gratz, “Hardware trojan threats to cache coherence in modern 2.5d chiplet systems,”IEEE Computer Architecture Letters, 2022

    2022

  16. [24]

    Secure interposer-based heterogeneous integration,

    M. Shafkat, M. Khan, C. Xi, A. A. Khan, M. T. Rahman, M. M. Tehra- nipoor, and N. Asadizanjani, “Secure interposer-based heterogeneous integration,”IEEE Design & Test, vol. 39, no. 6, pp. 156–164, 2022

    2022

  17. [25]

    On hardware security and trust for chiplet-based 2.5D and 3D ICs: Challenges and innovations,

    J. Suzano Da Fonseca, F. Abouzeid, G. Di Natale, A. Philippe, and P. Roche, “On hardware security and trust for chiplet-based 2.5D and 3D ICs: Challenges and innovations,”IEEE Access, vol. 12, pp. 29 778– 29 794, 2024

    2024

  18. [26]

    Security assurance for system-on-chip designs with untrusted ips,

    A. Basak, S. Bhunia, T. Tkacik, and S. Ray, “Security assurance for system-on-chip designs with untrusted ips,”IEEE Transactions on Information F orensics and Security, vol. 12, no. 7, pp. 1515–1528, 2017

    2017

  19. [27]

    Fort-NoCs: Mitigating the threat of a compromised NoC,

    D. M. Ancajas, K. Chakraborty, and S. Roy, “Fort-NoCs: Mitigating the threat of a compromised NoC,” inProc. DAC. ACM, 2014, pp. 158:1–158:6

    2014

  20. [28]

    Secure memory accesses on networks-on-chip,

    L. Fiorin, G. Palermo, S. Lukovic, V . Catalano, and C. Silvano, “Secure memory accesses on networks-on-chip,”IEEE Transactions on Comput- ers, vol. 57, no. 9, pp. 1216–1229, 2008

    2008

  21. [29]

    An interposer-based root of trust: Seize the opportunity for secure system-level integration of untrusted chiplets,

    M. Nabeel, M. Ashraf, S. Patnaik, V . Soteriou, O. Sinanoglu, and J. Knechtel, “An interposer-based root of trust: Seize the opportunity for secure system-level integration of untrusted chiplets,”CoRR, vol. abs/1906.02044, 2019. [Online]. Available: http://arxiv.org/abs/1906. 02044

    1906 arXiv

  22. [30]

    Cache hierarchy and memory subsystem of the amd opteron processor,

    P. Conway, N. Kalyanasundharam, G. Donley, K. Lepak, and B. Hughes, “Cache hierarchy and memory subsystem of the amd opteron processor,”IEEE Micro, vol. 30, no. 2, p. 16–29, 2010. [Online]. Available: https://doi.org/10.1109/MM.2010.31

    2010 doi

  23. [31]

    Concerted wire lifting: Enabling secure and cost-effective split manufacturing,

    S. Patnaik, M. Ashraf, H. Li, J. Knechtel, and O. Sinanoglu, “Concerted wire lifting: Enabling secure and cost-effective split manufacturing,” IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, vol. 41, no. 2, pp. 266–280, 2022

    2022

  24. [32]

    Rethinking split manufacturing: An information-theoretic approach with secure layout techniques,

    A. Sengupta, S. Patnaik, J. Knechtel, M. Ashraf, S. Garg, and O. Sinanoglu, “Rethinking split manufacturing: An information-theoretic approach with secure layout techniques,” inProc. ICCAD, 2017, pp. 329–336

    2017

  25. [33]

    Raise your game for split manufacturing: restoring the true functionality through beol,

    S. Patnaik, M. Ashraf, J. Knechtel, and O. Sinanoglu, “Raise your game for split manufacturing: restoring the true functionality through beol,” inProc. DAC, 2018. [Online]. Available: https: //doi.org/10.1145/3195970.3196100

    2018 doi

  26. [34]

    Attacking split manufacturing from a deep learning perspective,

    H. Li, S. Patnaik, A. Sengupta, H. Yang, J. Knechtel, B. Yu, E. F. Young, and O. Sinanoglu, “Attacking split manufacturing from a deep learning perspective,” inProc. DAC, 2019. [Online]. Available: https://doi.org/10.1145/3316781.3317780

    2019 doi

  27. [35]

    A new paradigm in split manufacturing: Lock the feol, unlock at the beol,

    A. Sengupta, M. Nabeel, M. Ashraf, J. Knechtel, and O. Sinanoglu, “A new paradigm in split manufacturing: Lock the feol, unlock at the beol,”Cryptography, vol. 6, no. 2, 2022. [Online]. Available: https://www.mdpi.com/2410-387X/6/2/22

    2022

  28. [36]

    Un-split: Attacking split manufacturing using link pre- diction in graph neural networks,

    L. Alrahis, L. Mankali, S. Patnaik, A. Sengupta, J. Knechtel, and O. Sinanoglu, “Un-split: Attacking split manufacturing using link pre- diction in graph neural networks,” inSecurity, Privacy, and Applied Cryptography Engineering, 2024, pp. 197–213

    2024

  29. [37]

    Best of both worlds: Integration of split manufacturing and camouflaging into a security-driven cad flow for 3d ics,

    S. Patnaik, M. Ashraf, O. Sinanoglu, and J. Knechtel, “Best of both worlds: Integration of split manufacturing and camouflaging into a security-driven cad flow for 3d ics,” inProc. ICCAD, 2018, pp. 1–8

    2018

  30. [38]

    A modern approach to ip protection and trojan prevention: Split manufacturing for 3d ics and obfuscation of vertical interconnects,

    ——, “A modern approach to ip protection and trojan prevention: Split manufacturing for 3d ics and obfuscation of vertical interconnects,” IEEE Transactions on Emerging Topics in Computing, vol. 9, no. 4, pp. 1815–1834, 2019

    2019

  31. [39]

    Securing computer hardware using 3d integrated circuit (ic) technology and split manufac- turing for obfuscation,

    F. Imeson, A. Emtenan, S. Garg, and M. Tripunitara, “Securing computer hardware using 3d integrated circuit (ic) technology and split manufac- turing for obfuscation,” inProc. USENIX Security, 2013, pp. 495–510

    2013

  32. [40]

    Security-aware 2.5 d integrated circuit design flow against hardware ip piracy,

    Y . Xie, C. Bao, and A. Srivastava, “Security-aware 2.5 d integrated circuit design flow against hardware ip piracy,”Computer, vol. 50, no. 5, pp. 62–71, 2017

    2017

  33. [41]

    Design flow for active interposer-based 2.5- d ics and study of risc-v architecture with secure noc,

    H. Park, J. Kim, V . C. K. Chekuri, M. A. Dolatsara, M. Nabeel, A. Bo- jesomo, S. Patnaik, O. Sinanoglu, M. Swaminathan, S. Mukhopadhyay, J. Knechtel, and S. K. Lim, “Design flow for active interposer-based 2.5- d ics and study of risc-v architecture with secure noc,”IEEE Tran...

    2020

Pith tools

Reviewed June 26, 2026 · model on record in the stance chip above.