Pith. sign in

REVIEW 3 major objections 2 minor 29 references

AuthenTree: A Scalable MPC-Based Distributed Trust Architecture for Chiplet-based Heterogeneous Systems

T0 review · 3 major / 2 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read The paper claims that chiplet authentication in multi-vendor system-in-package assemblies can be made distributed and trust-free using tree-structured multi-party computation, with negligible overhead.

desk verdict Plausible and potentially important chiplet-authentication architecture, but with the supplied full text unreadable, all substantive claims are unverifiable. read the letter →

arxiv 2508.13033 v1 pith:WONFE3ZJ submitted 2025-08-18 cs.CR

Ishraq Tashdid , Tasnuva Farheen , Sazadur Rahman This is my paper
classification cs.CR
keywords chipletauthenticationmulti-partycomputationsystem-in-package(SiP)distributedtrustzero-trusthardwareheterogeneousintegrationsupply-chainsecurityoverhead
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

AuthenTree tries to show that chiplet authentication in multi-vendor system-in-package assembly can be secured without a trusted integrator, a dedicated security chip, or any single point of failure. Instead, several ordinary integrator chiplets jointly perform the verification using multi-party computation arranged in a tree, so no one chiplet ever sees the raw signature. The authors report that, across five SiP benchmarks, the scheme costs as little as 0.48% area (about 7,000 square micrometers), under 0.5% power overhead, and less than one microsecond of authentication latency, in some cases several hundred times cheaper than previous approaches. If these numbers hold, the paper would remove the main practical obstacle to zero-trust authentication in chiplet-based systems.

What carries the argument

The load-bearing mechanism is tree-structured multi-party computation (MPC) over signature shares: a cryptographic protocol in which several parties jointly compute a function without revealing their private inputs. Each authentication request is answered by a set of integrator chiplets that collectively hold shares of the cryptographic material; no single chiplet can see the raw signature or decide the outcome alone. The tree topology limits how many partners each chiplet must talk to and lets partial results be aggregated level by level, which is what keeps the protocol scalable as the number of chiplets grows.

What would settle it

On a real or simulated SiP, corrupt any set of integrator chiplets up to the unstated threshold and attempt to reconstruct the signature from their MPC transcripts; if the raw signature or a valid forgery is recoverable from fewer than the intended number of parties, the central security claim is false. A second, cheaper check: synthesize the AuthenTree MPC blocks on a 28nm or 7nm standard-cell library and compare measured area and power against the claimed 0.48% and 0.5% bounds on the same five benchmarks.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the verification of a chiplet's digital signature can be split across multiple integrator chiplets such that the signature itself is never reconstructed, while the result of the check is still trustworthy. AuthenTree organizes the participating chiplets into a tree; a parent chiplet combines partial authentication results from its children rather than receiving the underlying secret. The architecture therefore distributes trust over the integrators already present in the system, eliminating the need for a centralized security anchor or dedicated hardware. The evaluation claims this is achieved with negligible area, power, and latency co

Load-bearing premise

The load-bearing premise is that the chiplets doing the multi-party computation will not collude beyond some unstated threshold; if enough of them pool their shares, they could reconstruct raw signatures or forge an authentication, so the entire "zero-trust" claim rests on that unstated limit.

Editorial extensions

If this is right

  • Chiplet vendors and system integrators could authenticate parts without exposing signing keys to any single assembler, reducing the risk of cloning and overproduction.
  • A system-in-package can maintain a zero-trust posture in which no individual integrator chiplet is a trusted third party whose compromise breaks the entire system.
  • The reported area, power, and latency figures put distributed authentication within reach of production SiP designs rather than relegating it to research-only overhead.
  • Authentication latency below one microsecond means the scheme can run at boot time or during operation without introducing a noticeable stall.
  • The tree architecture scales across the five evaluated SiP benchmarks, suggesting it can accommodate different package sizes and chiplet counts.

Reading between the lines

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

  • My inference: the paper's security claim should be read with an explicit corruption threshold; a natural next step would be to state and prove the protocol under a malicious-adversary model, since the abstract does not specify whether the MPC is semi-honest or actively secure.
  • An extension the authors do not pursue: the same tree-MPC pattern could be applied beyond chiplet dies, for example to authenticate modules, boards, or firmware updates in any multi-vendor assembly where no single party is trusted.
  • A concrete testable extension would be to synthesize the AuthenTree MPC logic on a mature process node and measure whether the 0.48% area and 0.5% power bounds hold at scale, varying the tree branching factor to map the latency/overhead trade-off.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 2 minor

Summary. AuthenTree proposes a tree-structured multi-party computation (MPC) framework for authenticating chiplets in heterogeneous system-in-package (SiP) assemblies. It claims to validate chiplets without revealing raw signatures and without a centralized trust anchor, distributing trust among integrator chiplets. The abstract reports evaluation on five SiP benchmarks with area overhead as low as 0.48% (7,000 sq-µm), power overhead under 0.5%, latency below 1 µs, and a speedup over prior work of up to 700× in some cases. The supplied full text is largely illegible (mojibake) and contains an unrelated astro-ph running head, so the protocol description, threat model, proofs, and experimental methodology cannot be audited in the reviewed record.

Significance. If the claims hold, AuthenTree addresses a genuinely important problem: chiplet authentication in multi-vendor supply chains currently relies on trusted integrators or centralized security anchors, and a scalable MPC-based alternative with negligible overhead would be a meaningful contribution to hardware and supply-chain security. The reported quantitative results—0.48% area, sub-µs latency, and large speedups—are the kind of concrete evidence that would matter to the chiplet design community. However, as presented, the manuscript provides no accessible derivation, no explicit adversary model, no baseline definition, and no reproducible evaluation setup. The potential significance is high, but the reviewed text does not currently support it.

major comments (3)
  1. [Abstract (security model)] The central claim—"secure chiplet validation without revealing raw signatures, distributing trust across multiple integrator chiplets"—is an MPC claim, but the abstract gives no corruption threshold t, no adversary type (semi-honest, malicious, or covert), no robustness/liveness assumption, and no statement of what a colluding subset of integrator chiplets can and cannot do. The phrase "zero-trust SiP environments" is therefore not justified: every MPC guarantee is conditional on a named threshold, and if t is small or the protocol is only semi-honest, a colluding subset could pool shares and reconstruct raw signatures or forge authentications. This is load-bearing and must be specified and proved.
  2. [Abstract (evaluation)] The headline evaluation numbers (0.48% area, <0.5% power, <1 µs latency, and 700× speedup) are stated without methodology: no synthesis or measurement setup, no process node, no benchmark definitions, no identification of the baseline for the 700× comparison, and no error bars or corner analysis. The unreadable body contains no legible equations, tables, or figures that could support these figures. As stated, the quantitative claims outrun the evidence and cannot be checked by a reviewer.
  3. [Full text (document integrity)] The supplied body text is not readable: most content is garbled characters, and the running head reads "arXiv:2508.13031v1 [astro-ph.GA] 18 Aug 2025," not the manuscript's own identifier. There is no legible protocol description, adversary model, correctness proof, complexity analysis, or experimental section. A journal submission must present a coherent audit trail; this text does not, so the central derivation is absent from the reviewed record. This is a reviewability defect, not a mere stylistic issue.
minor comments (2)
  1. [Abstract/threat model] If a corrected revision is supplied, add an explicit "Adversary Model and Trust Assumptions" subsection early in the paper, stating the corruption threshold t, the adversary type, and the guarantees that hold when shares are compromised.
  2. [Abstract/evaluation] Define "zero trust" operationally in a threshold-MPC setting, and spell out the baseline for the 700× comparison; per-benchmark tables with confidence intervals should accompany any such speedup claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity demonstrable: the supplied text is corrupted mojibake, so no equation or derivation chain can be audited.

full rationale

The full text supplied is unreadable: it consists almost entirely of mojibake, with an interleaved running head from an unrelated astro-ph paper ('arXiv:2508.13031v1 [astro-ph.GA] 18 Aug 2025'). No equations, algorithms, threat-model statements, or evaluation procedures are legible. The only readable portion is the abstract, which asserts that AuthenTree uses MPC in a tree architecture and reports area, power, and latency overheads. Under the hard rules, circularity requires quoting a specific reduction — e.g., a parameter fitted to a target quantity and then renamed as a prediction, or a definition that already contains the claimed result. None of that can be exhibited from the corrupted text. The absence of a stated adversary model or threshold is a security-model completeness concern, not a circularity concern. There is no visible load-bearing self-citation and no fitted-input-called-prediction step. The honest finding is therefore no significant circularity, with score 0.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

No free parameters are identifiable from the abstract alone. The axioms are the unstated security and implementation premises that the central claim depends on; each is a domain assumption because the abstract does not provide an adversary model, protocol specification, or overhead breakdown.

assumptions (3)
  • domain assumption A threshold of cooperating integrator chiplets cannot collude to forge or leak signatures.
    All MPC schemes require an explicit adversary and collusion threshold. The abstract claims 'distributing trust' and 'zero-trust' but states no threshold, so the core security guarantee is assumed.
  • domain assumption Chiplet authentication can be expressed as a secure joint verification of raw signatures under MPC with the claimed latency.
    The scheme's correctness and latency assume that signature verification can be evaluated distributively within 1 microsecond; the abstract gives no protocol details to support this.
  • domain assumption Inter-chiplet communication links inside the SiP can carry the MPC rounds without breaking the overhead budget.
    The overhead figures must include MPC message traffic between chiplets. The abstract does not show whether the area and power numbers account for the inter-chiplet communication fabric.

how reviews work

0 comments
Cite this review

Pith. "Pith review of AuthenTree: A Scalable MPC-Based Distributed Trust Architecture for Chiplet-based Heterogeneous Systems." pith.science (2026). https://pith.science/paper/WONFE3ZJ

@misc{pith2026250813033,
  author       = {Pith},
  title        = {Pith review of: AuthenTree: A Scalable MPC-Based Distributed Trust Architecture for Chiplet-based Heterogeneous Systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WONFE3ZJ}},
  note         = {Machine review of arXiv:2508.13033}
}
read the original abstract

The rapid adoption of chiplet-based heterogeneous integration is reshaping semiconductor design by enabling modular, scalable, and faster time-to-market solutions for AI and high-performance computing. However, multi-vendor assembly in post-fabrication environments fragments the supply chain and exposes SiP systems to serious security threats, including cloning, overproduction, and chiplet substitution. Existing authentication solutions depend on trusted integrators or centralized security anchors, which can expose sensitive data or create single points of failure. We introduce AuthenTree, a distributed authentication framework that leverages multi-party computation (MPC) in a scalable tree-based architecture, removing the need for dedicated security hardware or centralized trust. AuthenTree enables secure chiplet validation without revealing raw signatures, distributing trust across multiple integrator chiplets. Our evaluation in five SiP benchmarks demonstrates that AuthenTree imposes minimal overhead, with an area as low as 0.48% (7,000 sq-micrometers), an overhead power under 0.5%, and an authentication latency below 1 microsecond, surpassing previous work in some cases by 700 times. These results establish AuthenTree as an efficient, robust, and scalable solution for next-generation chiplet-based security in zero-trust SiP environments.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

29 extracted references · 28 canonical work pages

  1. [1]

    N. V. et al., `` ToSHI - towards secure heterogeneous integration: Security risks, threat assessment, and assurance,'' Cryptology ePrint Archive, Paper 2022/984, 2022. [Online]. Available: https://eprint.iacr.org/2022/984

    work page 2022

  2. [2]

    Sheikh et al., ``2.5d and 3d heterogeneous integration: Emerging applications,'' IEEE Solid-State Circuits Magazine, vol

    F. Sheikh et al., ``2.5d and 3d heterogeneous integration: Emerging applications,'' IEEE Solid-State Circuits Magazine, vol. 13, no. 4, pp. 77--87, 2021

    work page 2021

  3. [3]

    Tashdid, V

    I. Tashdid, V. Terry, J. Merkel, T. Farheen, and S. Rahman, ``Beyond PPA : Human-inspired reinforcement learning for post-route reliability aware macro placement,'' in International Symposium on Machine Learning for CAD, 2025

    work page 2025

  4. [4]

    J. H. Lau, Heterogeneous integrations. 1em plus 0.5em minus 0.4em Springer, 2019

    work page 2019

  5. [5]

    ECOLogic: Enabling Circular, Obfuscated, and Adaptive Logic via eFPGA-Augmented SoCs

    I. Tashdid, D. Saiham, N. Anjum, T. Farheen, and S. Rahman, ``Ecologic: Enabling circular, obfuscated, and adaptive logic via efpga-augmented socs,'' arXiv preprint arXiv:2508.04516, 2025

    work page Pith review arXiv 2025

  6. [6]

    Yahyaei, M

    K. Yahyaei, M. S. M. Khan, S. R. Larmann, N. Varshney, A. Bhattacharya, P. S. Shah, B. Chatterjee, and N. Asadizanjani, ``Ai trustworthiness in the era of advanced packaging: Challenges and opportunities,'' in 2025 IEEE 75th Electronic Components and Technology Conference (ECTC). 1em plus 0.5em minus 0.4em IEEE, 2025, pp. 911--919

    work page 2025

  7. [7]

    G. I. Haidar et al, ``Gate-sip: Enabling authenticated encryption testing in systems-in-package,'' in Proceedings of the 61st ACM/IEEE Design Automation Conference, ser. DAC '24. 1em plus 0.5em minus 0.4em New York, NY, USA: Association for Computing Machinery, 2024. [Online]. Available: https://doi.org/10.1145/3649329.3656527

    work page arXiv 2024

  8. [8]

    Chandra et al., ``A case study on ieee 1838 compliant multi-die 3dic dft implementation,'' in 2023 IEEE International Test Conference (ITC), 2023, pp

    A. Chandra et al., ``A case study on ieee 1838 compliant multi-die 3dic dft implementation,'' in 2023 IEEE International Test Conference (ITC), 2023, pp. 11--20

    work page 2023

Show all 29 references
  1. [9]

    M. S. Ul Islam Sami et al., ``Pqc-hi: Pqc-enabled chiplet authentication and key exchange in heterogeneous integration,'' in 2024 IEEE 74th Electronic Components and Technology Conference (ECTC), 2024, pp. 464--471

    2024

  2. [10]

    J. H. Lau, ``Chiplet heterogeneous integration,'' in Semiconductor Advanced Packaging. 1em plus 0.5em minus 0.4em Springer, 2021, pp. 413--439

    2021

  3. [11]

    M. T. Rahman et al., ``Defense-in-depth: A recipe for logic locking to prevail,'' Integration, vol. 72, pp. 39--57, 2020

    2020

  4. [12]

    Ibnat et al., ``Actiwate: Adaptive and design-agnostic active watermarking for ip ownership in modern socs,'' in 2023 60th ACM/IEEE Design Automation Conference (DAC), 2023, pp

    Z. Ibnat et al., ``Actiwate: Adaptive and design-agnostic active watermarking for ip ownership in modern socs,'' in 2023 60th ACM/IEEE Design Automation Conference (DAC), 2023, pp. 1--6

    2023

  5. [13]

    Alkabani, F

    Y. Alkabani, F. Koushanfar et al., ``Active hardware metering for intellectual property protection and security.'' in USENIX security symposium, vol. 20, 2007, pp. 1--20

    2007

  6. [14]

    G. I. Haidar et al., ``Sect-hi: Enabling secure testing for heterogeneous integration to prevent sip counterfeits.''

  7. [15]

    Tashdid, T

    I. Tashdid, T. Farheen, and S. Rahman, ``Safe-sip: Secure authentication framework for system-in-package using multi-party computation,'' in Proceedings of the Great Lakes Symposium on VLSI 2025, ser. GLSVLSI '25. 1em plus 0.5em minus 0.4em New York, NY, USA: Association for C...

    2025

  8. [16]

    E. M. Songhori, S. U. Hussain, A.-R. Sadeghi, T. Schneider, and F. Koushanfar, ``Tinygarble: Highly compressed and scalable sequential garbled circuits,'' in 2015 IEEE Symposium on Security and Privacy, 2015, pp. 411--428

    2015

  9. [17]

    M. S. Riazi, M. Javaheripi, S. U. Hussain, and F. Koushanfar, ``Mpcircuits: Optimized circuit generation for secure multi-party computation,'' in 2019 IEEE International Symposium on Hardware Oriented Security and Trust (HOST), 2019, pp. 198--207

    2019

  10. [18]

    Hashemi, S

    M. Hashemi, S. Tajik, and F. Ganji, ``Garblet: Multi-party computation for protecting chiplet-based systems,'' in 2025 IEEE 43rd VLSI Test Symposium (VTS). 1em plus 0.5em minus 0.4em IEEE, 2025, pp. 1--7

    2025

  11. [19]

    M. L. Rahman, A. M. Shuvo, J. Zhou, M. Tehranipoor, and F. Farahmandi, ``Sipmeter: Active hardware metering for system-in-package in heterogeneous integration,'' in 2025 IEEE 75th Electronic Components and Technology Conference (ECTC). 1em plus 0.5em minus 0.4em IEEE, 2025, pp...

    2025

  12. [20]

    Safari, P

    Y. Safari, P. Aghanoury, S. S. Iyer, N. Sehatbakhsh, and B. Vaisband, ``Hybrid obfuscation of chiplet-based systems,'' in 2023 60th ACM/IEEE Design Automation Conference (DAC), 2023, pp. 1--6

    2023

  13. [21]

    Rahman et al, ``Lle: Mitigating ic piracy and reverse engineering by last level edit,'' in International Symposium for Testing and Failure Analysis, vol

    S. Rahman et al, ``Lle: Mitigating ic piracy and reverse engineering by last level edit,'' in International Symposium for Testing and Failure Analysis, vol. 84741. 1em plus 0.5em minus 0.4em ASM International, 2023, pp. 360--369

    2023

  14. [22]

    Zheng, A

    Y. Zheng, A. R. Krishna, and S. Bhunia, ``Scanpuf: Robust ultralow-overhead puf using scan chain,'' in 2013 18th Asia and South Pacific Design Automation Conference (ASP-DAC). 1em plus 0.5em minus 0.4em IEEE, 2013, pp. 626--631

    2013

  15. [23]

    Li and D

    Z. Li and D. Wentzlaff, ``Lucie: A universal chiplet-interposer design framework for plug-and-play integration,'' in 2024 57th IEEE/ACM International Symposium on Microarchitecture (MICRO). 1em plus 0.5em minus 0.4em IEEE, 2024, pp. 423--436

    2024

  16. [24]

    Group, ``Cva6 risc-v cpu,'' 2024, accessed: November 2024

    O. Group, ``Cva6 risc-v cpu,'' 2024, accessed: November 2024. [Online]. Available: https://github.com/openhwgroup/cva6

    2024

  17. [25]

    Corporation, ``Nvdla hardware: Open-source deep learning accelerator,'' https://github.com/nvdla/hw, 2017, accessed: 2025-07-18

    N. Corporation, ``Nvdla hardware: Open-source deep learning accelerator,'' https://github.com/nvdla/hw, 2017, accessed: 2025-07-18

    2017

  18. [26]

    [Online]

    UltraEmbedded, ``Risc-v soc,'' 2024, accessed: November 2024. [Online]. Available: https://github.com/ultraembedded/riscv\_soc

    2024

  19. [27]

    Available: https://github.com/openhwgroup/cva6

    `` CVA6 RISC-V CPU .'' [Online]. Available: https://github.com/openhwgroup/cva6

  20. [28]

    Available: https://opencores.org/projects/ or1200\_soc

    ``or1200\_soc.'' [Online]. Available: https://opencores.org/projects/ or1200\_soc

  21. [29]

    A. C.-C. Yao, ``How to generate and exchange secrets,'' in 27th Annual Symposium on Foundations of Computer Science (sfcs 1986), 1986, pp. 162--167

    1986

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.