Pith. sign in

REVIEW 3 major objections 3 minor 17 references

An Analysis of HPC and Edge Architectures in the Cloud

T0 review · 3 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A study of 396 real AWS architectures maps how industry builds HPC and edge systems in the cloud.

desk verdict The paper can't be reviewed because the submitted full text is a different manuscript; the abstract describes a useful but unverifiable descriptive analysis of 396 AWS architectures. read the letter →

arxiv 2508.01494 v1 pith:4RDVIZ4G submitted 2025-08-02 cs.DC cs.SE

Steven Santillan , Cristina L. Abad This is my paper · ORCID
classification cs.DCcs.SE
keywords HPCincloudedgecomputingAWSarchitecturescontinuumstoragesystemsmachinelearningservicesarchitecturalcomplexityindustrypractice
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

This paper analyzes a dataset of 396 real-world cloud architectures deployed on AWS to find which ones contain high-performance computing (HPC) or edge components, and then characterizes how those systems are designed. It focuses on the prevalence and interplay of AWS services, the types of storage systems used, overall architectural complexity, and the use of machine learning services. If these characterizations hold, they provide a grounded picture of current industry practice for building HPC and edge solutions in the cloud continuum, which could guide both practitioners and future research.

What carries the argument

The central object is the dataset of 396 cloud architectures, which the paper uses as its empirical base. The method involves labeling each architecture for the presence of HPC or edge components, then measuring the prevalence of AWS services, the kinds of storage systems deployed, architectural complexity, and the use of machine learning services. These measurements together form the characterization that carries the argument.

What would settle it

A concrete test would be to independently obtain the component labels or deployment manifests for a random subset of the 396 architectures and compare them against the paper's HPC/edge identification. If the prevalence of HPC or edge components, or the service-mix statistics, differ substantially from the paper's reported numbers, the characterization would be called into question.

Watch

Extended reading notes

Core claim

The paper's central claim is that, by examining 396 real-world AWS architectures, it can identify which are HPC or edge oriented and describe their designs in terms of service prevalence, storage choices, complexity, and machine learning adoption. The authors argue that this characterization reveals recognizable patterns in how industry assembles robust, scalable HPC and edge systems on AWS, such as which services recur, how storage is layered, and how machine learning is being integrated into these architectures. The discovery is descriptive rather than prescriptive, but the paper frames it as a valuable snapshot of industry practice in the cloud continuum.

Load-bearing premise

The analysis assumes the 396-architecture dataset is representative of HPC and edge architectures in the cloud and that its component labels for identifying HPC or edge elements are accurate.

Editorial extensions

If this is right

  • Cloud architects gain a benchmark of which AWS services and storage patterns are common in real HPC and edge deployments, helping them choose sensible defaults.
  • Researchers can target gaps revealed by the characterization, such as underused machine learning services or storage types that appear rarely in practice.
  • Complexity findings could inform tooling and best practices for designing, deploying, and managing HPC and edge architectures on AWS.
  • If machine learning usage is widespread in these architectures, it suggests that ML is becoming a standard component of HPC and edge workflows in the cloud.
  • The 'cloud continuum' framing implies that HPC and edge are increasingly blending, which may push cloud providers toward more unified service offerings.

Reading between the lines

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

  • Because the dataset is drawn from AWS-published architectures, the sample may overrepresent well-architected or vendor-aligned designs; replicating the analysis on architectures from other cloud providers would test whether these patterns generalize.
  • The complexity and service-mix measurements could be cross-referenced with operational outcomes such as cost or reliability to infer which design patterns actually perform best, an extension the paper does not attempt.
  • The labeling methodology used to identify HPC or edge components could be applied to other architecture repositories to build a comparative landscape of cloud HPC and edge practices.
  • If the dataset's component labels are inaccurate, the prevalence and complexity claims could shift materially, so independent validation against original deployment manifests would strengthen the conclusions.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 3 minor

Summary. The paper (arXiv:2508.01494) is an empirical study that analyzes a dataset of 396 real-world AWS cloud architectures, identifying those with HPC or edge components and characterizing their use of AWS services, storage systems, architectural complexity, and machine-learning services. The abstract promises insights into current industry practice in the cloud continuum. However, the full text supplied for review is not this paper; it is arXiv:2508.01491, an unrelated manuscript on the homogenizing effects of large language models. Consequently, the only assessable content is the abstract, and none of the methodological details, dataset description, labeling criteria, or statistical analyses are available for evaluation.

Significance. If the results hold, the paper would provide a useful descriptive snapshot of how real-world AWS architectures incorporate HPC and edge components, with potential value for practitioners and researchers mapping the cloud continuum. The research question is timely and the use of a public dataset is a commendable starting point. However, the significance cannot be assessed from the submitted materials because the supporting methods and results are absent. The paper's contribution would rest entirely on the accuracy and representativeness of the component labels and on the dataset's provenance, neither of which is described in the abstract. No code, data, or validation artifacts are present in the submission package.

major comments (3)
  1. [Full text (supplied as arXiv:2508.01491)] The full text provided for this submission is a different paper, an unrelated manuscript on the homogenizing effect of large language models. This is not a minor formatting issue: it means that the methods, dataset description, labeling procedure, validation, and results needed to evaluate the central empirical claim are entirely missing. Every quantitative claim in the abstract (prevalence of AWS services, storage types, complexity, ML-service use) depends on the identification of which architectures contain HPC or edge components, and that identification procedure is not available for review. The manuscript as submitted cannot be verified or falsified.
  2. [Abstract, 'identify those architectures that contain HPC or edge components'] The abstract does not state the criteria for classifying an architecture as containing HPC or edge components, nor how the classification was performed and validated. If the labels were based on keyword matching, manual inspection, or self-reported tags, the prevalence figures and downstream comparisons could be substantially affected by false positives or false negatives. Without a description of the labeling protocol and its reliability, the paper's central descriptive claims are unfalsifiable from the available text.
  3. [Abstract, 'how representative these results are'] The abstract claims to discuss representativeness, but it does not describe the dataset's sampling frame. A dataset of 396 architectures drawn from public AWS case studies, if that is the source, would likely overrepresent successful, large-scale, or vendor-selected examples and exclude private, on-premises, or hybrid systems. The abstract gives no information on inclusion criteria, data collection, or how the dataset was originally assembled, all of which are necessary to assess generalizability. This is a load-bearing gap because the paper's stated contribution is an empirical characterization of industry practice.
minor comments (3)
  1. [Abstract] The abstract would benefit from a citation or link to the dataset being analyzed, along with a statement of its original source and publication venue, to allow readers to assess its provenance.
  2. [Abstract, 'HPC or edge components'] The terms 'HPC' and 'edge' are not defined in the abstract; providing a brief operational definition or a pointer to the definition in the full text would improve clarity.
  3. [Full text] The supplied full text belongs to a different paper and includes its own abstract, glossary, and acknowledgments; this should be corrected before any further review is attempted.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found; the paper is an empirical analysis of an external dataset with no fitted predictions or self-citation chain.

full rationale

The target paper (arXiv:2508.01494) claims to analyze a dataset of 396 real-world AWS cloud architectures, identify those with HPC or edge components, and characterize service prevalence, storage, complexity, and ML-service use. This is a descriptive empirical study rather than a derivation: no equation is derived, no parameter is fitted to a subset and then renamed as a prediction, and no load-bearing argument depends on a self-citation. The possible weaknesses, such as representativeness of the dataset or accuracy of component labels, are data-quality and generalization concerns, not circularity. The supplied full text is actually a different paper (arXiv:2508.01491) on LLM homogenization, so the methods of the target paper cannot be inspected; that is a missing-evidence condition, not evidence of circularity. Under the hard rule that circularity must be demonstrated by quoting a specific reduction, and none is available, the appropriate verdict is no significant circularity with score 0.

Assumptions & free parameters 0 free parameters · 2 assumptions · 0 invented entities

The central claims rest on two domain assumptions: dataset representativeness and label correctness. These cannot be checked without the full method section and the dataset itself.

assumptions (2)
  • domain assumption The dataset of 396 AWS architectures is representative of HPC and edge architectures in the cloud.
    The abstract draws general conclusions about HPC and edge architectures in the cloud from an AWS-only dataset, so representativeness is load-bearing.
  • domain assumption The published dataset's labels for identifying HPC and edge components are correct.
    The prevalence and complexity findings depend on these labels, and the abstract does not describe how they were validated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of An Analysis of HPC and Edge Architectures in the Cloud." pith.science (2026). https://pith.science/paper/4RDVIZ4G

@misc{pith2026250801494,
  author       = {Pith},
  title        = {Pith review of: An Analysis of HPC and Edge Architectures in the Cloud},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4RDVIZ4G}},
  note         = {Machine review of arXiv:2508.01494}
}
read the original abstract

We analyze a recently published dataset of 396 real-world cloud architectures deployed on AWS, from companies belonging to a wide range of industries. From this dataset, we identify those architectures that contain HPC or edge components and characterize their designs. Specifically, we investigate the prevalence and interplay of AWS services within these architectures, examine the types of storage systems employed, assess architectural complexity and the use of machine learning services, discuss the implications of our findings and how representative these results are of HPC and edge architectures in the cloud. This characterization provides valuable insights into current industry practices and trends in building robust and scalable HPC and edge solutions in the cloud continuum, and can be valuable for those seeking to better understand how these architectures are being built and to guide new research.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

17 extracted references · 17 canonical work pages

  1. [1]

    11em plus .33em minus .07em 4000 4000 100 4000 4000 500 `\.=1000 = #1 \@IEEEnotcompsoconly \@IEEEcompsoconly #1 * [1] 0pt [0pt][0pt] #1 * [1] 0pt [0pt][0pt] #1 * \| ** #1 \@IEEEauthorblockNstyle \@IEEEcompsocnotconfonly \@IEEEauthorblockAstyle \@IEEEcompsocnotconfonly \@IEEEcompsocconfonly \@IEEEauthordefaulttextstyle \@IEEEcompsocnotconfonly \@IEEEauthor...

    work page

  2. [2]

    AWS Documentation , `` AWS well-architected framework,'' https://docs.aws.amazon.com/wellarchitected/latest/framework/welcome.html, last accessed: June 20, 2025

    work page 2025

  3. [3]

    Satija, C

    S. Satija, C. Ye, R. Kosgi, A. Jain, R. Kankaria, Y. Chen, A. Arpaci-Dusseau, R. Arpaci-Dusseau, and Srinivasan, ``Cloudscape: A study of storage services in modern cloud architectures,'' in USENIX FAST, 2025

    work page 2025

  4. [4]

    AWS Documentation , ``High performance computing on AWS ,'' https://aws.amazon.com/hpc/, last accessed: June 3rd, 2025

    work page 2025

  5. [5]

    ------, `` AWS for the edge,'' https://aws.amazon.com/edge/services/, last accessed: June 3rd, 2025

    work page 2025

  6. [6]

    ------, ``High performance computing lens: Traditional cluster environment,'' https://docs.aws.amazon.com/wellarchitected/latest/high-performance-computing-lens/traditional-cluster-environment.html, last accessed: June 20, 2025

    work page 2025

  7. [7]

    Eismann, J

    S. Eismann, J. Scheuner, E. van Eyk, M. Schwinger, J. Grohmann, N. Herbst, C. L. Abad, and A. Iosup, ``Serverless applications: Why, when, and how?'' IEEE Software, vol. 38, no. 1, 2021

    work page 2021

  8. [8]

    Eismann, J

    S. Eismann, J. Scheuner, E. v. Eyk, M. Schwinger, J. Grohmann, N. Herbst, C. L. Abad, and A. Iosup, ``The state of serverless applications: Collection, characterization, and community consensus,'' IEEE Transactions on Software Engineering, vol. 48, no. 10, 2022

    work page 2022

Show all 17 references
  1. [9]

    Eskandani and G

    N. Eskandani and G. Salvaneschi, ``The W onderless dataset for serverless computing,'' in IEEE/ACM Intl. Conf. Mining Softw. Repo. (MSR), 2021

    2021

  2. [10]

    Chavez-Moreno and C

    A. Chavez-Moreno and C. Abad, `` OpenLambdaVerse : A dataset and analysis of open-source serverless applications,'' in IEEE Intl. Conference on Cloud Engineering (IC2E), 2025

    2025

  3. [11]

    Gustafsson, S

    O. Gustafsson, S. Wilkinson, F. Bacall, S. Soiland-Reyes, S. Leo, L. Pireddu, S. Owen, N. Juty et al., `` WorkflowHub : A registry for computational workflows,'' Scientific Data, vol. 12, no. 1, 2025

    2025

  4. [12]

    Moreno-Vozmediano, E

    R. Moreno-Vozmediano, E. Huedo, R. S. Montero, and I. M. Llorente, ``A disaggregated cloud architecture for edge computing,'' IEEE Internet Computing, vol. 23, no. 3, 2019

    2019

  5. [13]

    J. Li, C. Gu, Y. Xiang, and F. Li, ``Edge-cloud computing systems for smart grid: state-of-the-art, architecture, and applications,'' Journal of Modern Power Systems and Clean Energy, vol. 10, no. 4, 2022

    2022

  6. [14]

    Belcastro, C

    L. Belcastro, C. Cosentino, F. Marozzo et al., ``Empowering efficient drone monitoring with low-latency edge-cloud continuum platforms,'' in Euromicro Intl. Conf. Par., Distrib., and Network-Based Proc., 2025

    2025

  7. [15]

    Mateescu, W

    G. Mateescu, W. Gentzsch, and C. J. Ribbens, ``Hybrid computing—where HPC meets grid and cloud computing,'' Future Generation Computer Systems, vol. 27, no. 5, 2011

    2011

  8. [16]

    G. G. Casta \ n \'e , H. Xiong, D. Dong, and J. P. Morrison, ``An ontology for heterogeneous resources management interoperability and hpc in the cloud,'' Future Generation Computer Systems, vol. 88, 2018

    2018

  9. [17]

    Imbrosciano, E

    M. Imbrosciano, E. Sciacca, F. Vitello, L. Pelonero, F. Franchina, U. Becciani, I. Colonnelli, and D. Medic, `` The Cloud-HPC infrastructure for Hazard Mapping and vulnerability Monitoring (HaMMon) ,'' in Euromicro Intl. Conf. Par., Distrib., and Network-Based Proc., 2025

    2025

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.