pith. machine review for the scientific record. sign in

arxiv: 2605.01575 · v1 · submitted 2026-05-02 · 💻 cs.PF · cs.AR

Recognition: unknown

SPEC CPU: The Next Generation

Mahesh Madhav , Allen Lee , Andres Mejia , Branden Moore , Charan Soppadandi , Chris Cambly , Christoph M\"ullner , Daniel Bowers
show 26 more authors
David Reiner Denis Bakhvalov Di Zhao Duane Voth Feng Xue Fr\'ed\'erique Silber-Chaussumier James Bucek James Southern Jiangning Liu Jim Himer John Henning Kevin Smith Kristen Yang Kunal Kashyap Mason Guy Mat Colgrove Michael Berg Prasad Battini Prasad Joshi Rohit Prasad Shayantika Bhattacharya Sriyash Caculo Stefan Reimbold Sundar Iyengar Van Smith Zarko Todorovski
Authors on Pith no claims yet

Pith reviewed 2026-05-10 15:07 UTC · model grok-4.3

classification 💻 cs.PF cs.AR
keywords SPEC CPU 2026CPU benchmarksperformance evaluationmultiprogrammed workloadsbenchmarking methodologyopen-source applicationsRolling-Round-Robin Rate
0
0 comments X

The pith

SPEC CPU 2026 introduces Rolling-Round-Robin Rate to standardize benchmarking of heterogeneous multiprogrammed workloads.

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

The paper presents SPEC CPU 2026 as the next CPU benchmark suite developed through community collaboration on modern open-source applications. These applications were chosen and adapted to emphasize workload diversity, portability, and long-term software relevance. The suite expands multithreaded benchmarks and adds workloads with varied microarchitectural behaviors to match current software patterns. Its primary new contribution is Rolling-Round-Robin Rate, a standardized method for executing mixed multiprogrammed workloads that fills an existing gap in consistent benchmarking practice. A reader would care because this approach promises more representative processor performance measurements than prior suites.

Core claim

SPEC CPU 2026 is constructed from modern open-source applications selected and hardened via a principled process focused on workload diversity, portability, and software longevity. It includes an expanded collection of multithreaded benchmarks along with workloads that exhibit distinct microarchitectural profiles. The central innovation is Rolling-Round-Robin Rate, which supplies a novel and standardized way to run heterogeneous, multiprogrammed workloads and thereby addresses a long-standing gap in benchmarking practice.

What carries the argument

Rolling-Round-Robin Rate, a novel standardized approach to running heterogeneous, multiprogrammed workloads that addresses a long-standing gap in benchmarking practice.

If this is right

  • The suite enables performance comparisons that better match the demands of contemporary software.
  • Expanded multithreaded benchmarks support evaluation of parallel execution on modern processors.
  • Workloads with distinct microarchitectural profiles increase coverage of different processor behaviors.
  • The standardized rate method reduces inconsistencies in how heterogeneous workloads are measured across systems.

Where Pith is reading between the lines

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

  • Widespread use could encourage hardware designs that optimize specifically for mixed rather than uniform workloads.
  • Benchmark results may become more comparable across research groups and vendors.
  • The approach could be adapted to other benchmark suites that currently lack a standard for multiprogrammed execution.

Load-bearing premise

The selected modern open-source applications will maintain workload diversity, portability, and software longevity while reflecting the demands of contemporary software.

What would settle it

If repeated runs of SPEC CPU 2026 using Rolling-Round-Robin Rate produce performance rankings and variability identical to those obtained from prior SPEC CPU suites on the same hardware, the claim that the new method fills a benchmarking gap would be falsified.

Figures

Figures reproduced from arXiv: 2605.01575 by Allen Lee, Andres Mejia, Branden Moore, Charan Soppadandi, Chris Cambly, Christoph M\"ullner, Daniel Bowers, David Reiner, Denis Bakhvalov, Di Zhao, Duane Voth, Feng Xue, Fr\'ed\'erique Silber-Chaussumier, James Bucek, James Southern, Jiangning Liu, Jim Himer, John Henning, Kevin Smith, Kristen Yang, Kunal Kashyap, Mahesh Madhav, Mason Guy, Mat Colgrove, Michael Berg, Prasad Battini, Prasad Joshi, Rohit Prasad, Shayantika Bhattacharya, Sriyash Caculo, Stefan Reimbold, Sundar Iyengar, Van Smith, Zarko Todorovski.

Figure 1
Figure 1. Figure 1: Self-similarity recurrence plot alongside a performance plot collected [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Time plots showing all the integer rate benchmarks executing on schedules for homogenous (refrate) and heterogeneous (rrrrate) on 48 copies. In [PITH_FULL_IMAGE:figures/full_fig_p012_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: BBV Recurrence and Perf Plots: Integer Rate, and single-threaded Integer Speed [PITH_FULL_IMAGE:figures/full_fig_p021_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: BBV Recurrence and Perf Plots: Floating Point Rate [PITH_FULL_IMAGE:figures/full_fig_p022_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Perf Plots: Integer Speed 23 [PITH_FULL_IMAGE:figures/full_fig_p023_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Perf Plots: Floating Point Speed 24 [PITH_FULL_IMAGE:figures/full_fig_p024_6.png] view at source ↗
read the original abstract

The march toward developing relevant and robust CPU benchmarks continues with the introduction of SPEC CPU 2026, the next generation suite for measuring processor performance. This paper details the methodology behind its creation, showcasing a process centered on community collaboration and principled development. The suite is built upon a foundation of modern, open-source applications, selected and hardened through a process that emphasizes workload diversity, portability, and software longevity. A key contribution is Rolling-Round-Robin Rate, a novel and standardized approach to running heterogeneous, multiprogrammed workloads that addresses a long-standing gap in benchmarking practice. Additionally, the suite features an expanded set of multithreaded benchmarks and introduces workloads with distinct microarchitectural profiles, reflecting the demands of contemporary software. By detailing our principled approach to benchmark selection, adaptation, and validation, we demonstrate how the SPEC CPU 2026 suite sets the standard for performance evaluation in the next era of computer architecture research and development.

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

2 major / 1 minor

Summary. The paper describes the development methodology for the SPEC CPU 2026 benchmark suite. It centers on community-driven selection and hardening of modern open-source applications to ensure workload diversity, portability, and software longevity. A primary contribution is the introduction of Rolling-Round-Robin Rate, a standardized scheduling approach for heterogeneous multiprogrammed workloads. The suite also expands multithreaded benchmarks and incorporates workloads with distinct microarchitectural profiles to better reflect contemporary software demands.

Significance. If the described process and new rate-based method hold up under scrutiny, this work would provide the computer architecture community with a timely, standardized benchmark suite that addresses longstanding gaps in evaluating multiprogrammed and heterogeneous workloads. The community-collaborative, open-source foundation and focus on modern applications represent a constructive evolution of SPEC practices, potentially improving relevance for next-generation processor research.

major comments (2)
  1. [Abstract and methodology sections] The central claims regarding principled selection, hardening, and validation of the new suite (including workload diversity and portability) are presented as process descriptions without accompanying validation data, error-handling details, or empirical results. This is load-bearing for the assertion that SPEC CPU 2026 sets the standard for performance evaluation.
  2. [Section describing Rolling-Round-Robin Rate] Rolling-Round-Robin Rate is positioned as a novel solution to a benchmarking gap, yet the manuscript provides no quantitative comparison to prior scheduling approaches (e.g., standard round-robin or rate-based methods) or evidence of its standardization benefits across heterogeneous workloads.
minor comments (1)
  1. [Abstract] The abstract and introduction would benefit from clearer delineation between descriptive process elements and any quantitative or falsifiable claims to help readers assess the strength of the contributions.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their positive assessment of the potential impact of SPEC CPU 2026 and for the constructive feedback. We respond to each major comment below and outline the revisions we will make to the manuscript.

read point-by-point responses

  1. Referee: [Abstract and methodology sections] The central claims regarding principled selection, hardening, and validation of the new suite (including workload diversity and portability) are presented as process descriptions without accompanying validation data, error-handling details, or empirical results. This is load-bearing for the assertion that SPEC CPU 2026 sets the standard for performance evaluation.

    Authors: The manuscript focuses on detailing the methodology and community-driven process for developing the SPEC CPU 2026 suite, including how applications were selected, hardened, and validated for diversity, portability, and longevity. We acknowledge that explicit validation data, error-handling procedures, and empirical results would strengthen the presentation of these claims. In the revised version, we will add a dedicated subsection in the methodology section that includes quantitative metrics from the validation process, details on error handling during hardening, and empirical evidence supporting workload diversity and portability. revision: yes

  2. Referee: [Section describing Rolling-Round-Robin Rate] Rolling-Round-Robin Rate is positioned as a novel solution to a benchmarking gap, yet the manuscript provides no quantitative comparison to prior scheduling approaches (e.g., standard round-robin or rate-based methods) or evidence of its standardization benefits across heterogeneous workloads.

    Authors: We agree that the manuscript would benefit from quantitative comparisons to demonstrate the advantages of Rolling-Round-Robin Rate over existing approaches. The current text introduces the method as a standardized scheduling approach for heterogeneous multiprogrammed workloads. We will revise the relevant section to include a quantitative evaluation, such as performance metrics and fairness measures across sample heterogeneous workloads, comparing it to standard round-robin and other rate-based scheduling methods. This will provide evidence of its standardization benefits. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper is a descriptive engineering report detailing the community-driven development process for the SPEC CPU 2026 suite, including application selection criteria and the introduction of Rolling-Round-Robin Rate as a scheduling method for heterogeneous workloads. No equations, quantitative predictions, fitted parameters, or derivations are present that could reduce by construction to inputs, self-citations, or ansatzes. The central claims rest on process descriptions and methodological choices rather than any self-referential logic, making the work self-contained without circular steps.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The claim depends on assumptions about workload representativeness and the novelty of the rate method; no free parameters or invented physical entities are introduced.

axioms (1)
  • domain assumption Modern open-source applications provide sufficient workload diversity, portability, and longevity to serve as the foundation for future CPU benchmarks.
    Invoked in the selection and hardening process described in the abstract.
invented entities (1)
  • Rolling-Round-Robin Rate no independent evidence
    purpose: Novel standardized method for running heterogeneous multiprogrammed workloads
    Introduced as the key contribution to fill a benchmarking gap; no independent evidence provided.

pith-pipeline@v0.9.0 · 5590 in / 1235 out tokens · 89522 ms · 2026-05-10T15:07:06.772449+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

184 extracted references · 73 canonical work pages · 1 internal anchor

  1. [1]

    For Better or Worse, Benchmarks Shape a Field: Technical Perspective,

    D. Patterson, “For Better or Worse, Benchmarks Shape a Field: Technical Perspective,”Commun. ACM, vol. 55, no. 7, p. 104, Jul. 2012,ISSN: 0001-0782.DOI: 10.1145/2209249.2209271

    work page doi:10.1145/2209249.2209271 2012

  2. [2]

    Overview of the SPEC Benchmarks,

    K. M. Dixit, “Overview of the SPEC Benchmarks,” inThe Benchmark Handbook, 1993. [Online]. Available: https : / / api . semanticscholar.org/CorpusID:6617287

    1993

  3. [3]

    SPEC CPU2000: Measuring CPU Performance in the New Millennium,

    J. L. Henning, “SPEC CPU2000: Measuring CPU Performance in the New Millennium,”Computer, vol. 33, no. 7, 28–35, Jul. 2000, ISSN: 0018-9162.DOI: 10.1109/2.869367

    work page doi:10.1109/2.869367 2000

  4. [4]

    SPEC CPU2006 benchmark descriptions,

    J. L. Henning, “SPEC CPU2006 benchmark descriptions,” SIGARCH Comput. Archit. News, 2006.DOI: 10 . 1145 / 1186736 . 1186737

    2006

  5. [5]

    SPEC CPU2017: Next-Generation Compute Benchmark,

    J. Bucek, K.-D. Lange, and J. v. Kistowski, “SPEC CPU2017: Next-Generation Compute Benchmark,” inCompanion of the 2018 ACM/SPEC International Conference on Performance Engineer- ing, ser. ICPE ’18, Berlin, Germany: ACM, 2018, 41–42,ISBN: 9781450356299.DOI: 10.1145/3185768.3185771

    work page doi:10.1145/3185768.3185771 2018

  6. [6]

    Munafo,The SPEC Benchmarks, 2025

    R. Munafo,The SPEC Benchmarks, 2025. [Online]. Available: https: //www.mrob.com/pub/comp/benchmarks/spec.html

    2025

  7. [7]

    : Interactive dynamics for visual analysis

    A. Danowitz et al., “CPU DB: Recording Microprocessor History,” Commun. ACM, vol. 55, no. 4, 55–63, Apr. 2012,ISSN: 0001-0782. DOI: 10.1145/2133806.2133822

    work page doi:10.1145/2133806.2133822 2012

  8. [8]

    T˚ uma,Long Term Stability Observations, DaCapo Benchmark,

    P. T˚ uma,Long Term Stability Observations, DaCapo Benchmark,

  9. [9]

    Available: https : / / github

    [Online]. Available: https : / / github . com / dacapobench / dacapobench/issues/269

  10. [10]

    Kessler,Variance in DaCapo Benchmark, DaCapo Benchmark,

    P. Kessler,Variance in DaCapo Benchmark, DaCapo Benchmark,

  11. [11]

    Available: https : / / github

    [Online]. Available: https : / / github . com / dacapobench / dacapobench/issues/302#issuecomment-2469456922

  12. [12]

    A. M. Yang,Large Variance in Exec Time for scala-doku, Renais- sance Benchmark, 2019. [Online]. Available: https://github.com/ renaissance-benchmarks/renaissance/issues/175

    2019

  13. [13]

    Bulej,Renaissance 0.16, Renaissance Benchmark, 2024

    L. Bulej,Renaissance 0.16, Renaissance Benchmark, 2024. [On- line]. Available: https://renaissance.dev/2024/11/22/renaissance-0- 16-0.html

    2024

  14. [14]

    Many Benchmarks Stress the Same Bottlenecks,

    H. Vandierendonck and K. D. Bosschere, “Many Benchmarks Stress the Same Bottlenecks,” inWorkshop on Computer Architecture Evaluation Using Commercial Workloads, 2004. [Online]. Available: https://api.semanticscholar.org/CorpusID:11877774

    2004

  15. [15]

    Wang et al.,A Detailed Historical and Statistical Analysis of the Influence of Hardware Artifacts on SPEC Integer Benchmark Performance, 2024

    Y . Wang et al.,A Detailed Historical and Statistical Analysis of the Influence of Hardware Artifacts on SPEC Integer Benchmark Performance, 2024. arXiv: 2401.16690[cs.CY]

    work page arXiv 2024

  16. [16]

    Wang et al.,Achieving Consistent and Comparable CPU Evalu- ation, 2025

    C. Wang et al.,Achieving Consistent and Comparable CPU Evalu- ation, 2025. arXiv: 2411.08494[cs.PF]

    work page arXiv 2025

  17. [17]

    [Online]

    The Tukaani Project,XZ Utils. [Online]. Available: https : / / en . wikipedia.org/wiki/XZ_Utils

  18. [18]

    Romstad, M

    T. Romstad, M. Costalba, and J. Kiiski,Stockfish: Strong open- source chess engine. [Online]. Available: https://stockfishchess.org/

  19. [19]

    Welty and V

    C. Welty and V . Petric,NTest: A Strong Othello Program. [Online]. Available: https://github.com/vladpetric/ntest

  20. [20]

    Kiehn, M

    K. P. Gaffney et al., “SQLite: past, present, and future,”Proceedings of the VLDB Endowment, vol. 15, no. 12, 2022.DOI: 10.14778/ 3554821.3554842 [Online]. Available: https://sqlite.org/about.html

    work page arXiv 2022

  21. [21]

    An Overview of the OMNeT++ Simu- lation Environment,

    A. Varga and R. Hornig, “An Overview of the OMNeT++ Simu- lation Environment,” inProceedings of the 1st International Con- ference on Simulation Tools and Techniques for Communications, Networks and Systems & Workshops, ser. Simutools ’08, Marseille, France: ICST, 2008,ISBN: 9789639799202.DOI: 10.4108/ICST. SIMUTOOLS2008.3027 [Online]. Available: www.omnetpp.org

    work page doi:10.4108/icst 2008

  22. [22]

    Python reference manual,

    G. Rossum, “Python reference manual,” CWI (Centre for Mathemat- ics and Computer Science), Netherlands, Tech. Rep., 1995. [Online]. Available: https://ir.cwi.nl/pub/5008

    1995

  23. [23]

    Coalson, E

    J. Coalson, E. de Castro Lopo, and M. van Beurden,flac: Free Lossless Audio Codec. [Online]. Available: https://xiph.org/flac

  24. [24]

    Stallman,GCC, the GNU Compiler Collection

    R. Stallman,GCC, the GNU Compiler Collection. [Online]. Avail- able: https://gcc.gnu.org

  25. [25]

    LLVM: A Compilation Framework for Lifelong Program Analysis & Transformation

    C. Lattner and V . Adve, “LLVM: A Compilation Framework for Lifelong Program Analysis & Transformation,” inInternational Symposium on Code Generation and Optimization, 2004. CGO 2004., 2004, pp. 75–86.DOI: 10.1109/CGO.2004.1281665 [Online]. Available: www.llvm.org

    work page doi:10.1109/cgo.2004.1281665 2004

  26. [26]

    Marjamäki,Cppcheck: a tool for static C/C++ code analysis

    D. Marjamäki,Cppcheck: a tool for static C/C++ code analysis. [Online]. Available: www.cppcheck.com

  27. [27]

    ABC: An Academic Industrial- Strength Verification Tool,

    R. Brayton and A. Mishchenko, “ABC: An Academic Industrial- Strength Verification Tool,” inComputer Aided Verification, Berlin, Heidelberg: Springer Berlin Heidelberg, 2010, pp. 24–40,ISBN: 978-3-642-14295-6. [Online]. Available: https : / / people . eecs . berkeley.edu/~alanmi/abc/

    2010

  28. [28]

    VTR 9: Open-Source CAD for Fabric and Beyond FPGA Architecture Exploration,

    M. A. Elgammal et al., “VTR 9: Open-Source CAD for Fabric and Beyond FPGA Architecture Exploration,”ACM Trans. Reconfig- urable Technol. Syst., vol. 18, Aug. 2025,ISSN: 1936-7406.DOI: 10.1145/3734798 [Online]. Available: https://verilogtorouting.org

    work page doi:10.1145/3734798 2025

  29. [29]

    M., Akram, A., Alian, M., Amslinger, R., An- dreozzi, M., Armejach, A., Asmussen, N., Beckmann, B., Bharadwaj, S., et al

    J. Lowe-Power et al., “The gem5 Simulator: Version 20.0+,” 2020. arXiv: 2007.03152. [Online]. Available: www.gem5.org

    work page arXiv 2020

  30. [30]

    Sensitive protein align- ments at tree-of-life scale using DIAMOND,

    B. Buchfink, K. Reuter, and H.-G. Drost, “Sensitive protein align- ments at tree-of-life scale using DIAMOND,” inNature Methods, 13 vol. 18, 2021.DOI: 10 . 1038 / s41592 - 021 - 01101 - x [Online]. Available: https://github.com/bbuchfink/diamond

    2021

  31. [31]

    MiniZinc: Towards a Standard CP Modelling Language,

    N. Nethercote et al., “MiniZinc: Towards a Standard CP Modelling Language,” inPrinciples and Practice of Constraint Programming – CP 2007, Berlin, Heidelberg: Springer, 2007, pp. 529–543,ISBN: 978-3-540-74970-7. [Online]. Available: www.minizinc.org

    2007

  32. [32]

    Simple Encrypted Arithmetic Library - SEAL v2.1,

    H. Chen, K. Laine, and R. Player, “Simple Encrypted Arithmetic Library - SEAL v2.1,” inFinancial Cryptography and Data Security, Cham: Springer International Publishing, 2017, pp. 3–18,ISBN: 978- 3-319-70278-0.DOI: 10 . 1007 / 978 - 3 - 319 - 70278 - 0 _ 1 [Online]. Available: https://github.com/Microsoft/SEAL

    2017

  33. [33]

    The ns-3 Network Simulator,

    G. F. Riley and T. R. Henderson, “The ns-3 Network Simulator,” inModeling and Tools for Network Simulation. Berlin, Heidelberg: Springer, 2010, pp. 15–34,ISBN: 978-3-642-12331-3.DOI: 10.1007/ 978-3-642-12331-3_2 [Online]. Available: www.nsnam.org

    2010

  34. [34]

    Introducing the Graph 500,

    R. C. Murphy et al., “Introducing the Graph 500,”Cray Users Group (CUG), vol. 19, no. 45-74, p. 22, 2010. [Online]. Available: https: //graph500.org

    2010

  35. [35]

    [Online]

    ZStd Community,ZStd, the C reference implementation for the ZStandard lossless compression algorithm. [Online]. Available: https://facebook.github.io/zstd

  36. [36]

    A., Caplan, R

    C. Downs et al., “A Near-real-time Data-assimilative Model of the Solar Corona,”Science, vol. 388, no. 6753, pp. 1306–1310, 2025. DOI: 10.1126/science.adq0872

    work page doi:10.1126/science.adq0872 2025

  37. [37]

    Quadratic Gravity,

    D. Guerrera et al., “Towards a Mini-App for Smoothed Particle Hydrodynamics at Exascale,” in2018 IEEE International Confer- ence on Cluster Computing (CLUSTER), 2018, pp. 607–614.DOI: 10.1109/CLUSTER.2018.00077

    work page doi:10.1109/cluster.2018.00077 2018

  38. [38]

    The Cactus Framework and Toolkit: Design and Applications,

    T. Goodale et al., “The Cactus Framework and Toolkit: Design and Applications,” inVECPAVector and Parallel Processing R’2002, 5th International Conference, Springer (2003), vol. 2565, Jun. 2002, ISBN: 978-3-540-00852-1.DOI: 10 . 1007 / 3 - 540 - 36569 - 9 _ 13 [Online]. Available: www.cactuscode.org

    2002

  39. [39]

    TeaLeaf: A Mini-Application to Enable Design-Space Explorations for Iterative Sparse Linear Solvers,

    S. McIntosh-Smith et al., “TeaLeaf: A Mini-Application to Enable Design-Space Explorations for Iterative Sparse Linear Solvers,” in2017 IEEE International Conference on Cluster Computing (CLUSTER), 2017, pp. 842–849.DOI: 10.1109/CLUSTER.2017.105

    work page doi:10.1109/cluster.2017.105 2017

  40. [40]

    Modeling Unusual Nucleic Acid Structures,

    T. J. Macke and D. A. Case, “Modeling Unusual Nucleic Acid Structures,” inMolecular Modeling of Nucleic Acids. American Chemical Society, 1997, ch. 24, pp. 379–393.DOI: 10.1021/bk- 1998-0682.ch024 [Online]. Available: https://casegroup.rutgers.edu

    work page doi:10.1021/bk- 1997

  41. [41]

    Mantevo 3.0 Overview,

    M. A. Heroux, “Mantevo 3.0 Overview,” Sandia National Lab.(SNL- NM), Albuquerque, NM (United States), Tech. Rep., 2015. [Online]. Available: https://mantevo.github.io/applications.html

    2015

  42. [42]

    Overview of the PALM model system 6.0,

    B. Maronga et al., “Overview of the PALM model system 6.0,” Geoscientific Model Development, vol. 13, no. 3, pp. 1335–1372, 2020.DOI: 10.5194/gmd-13-1335-2020 [Online]. Available: https: //docs.palm-model.com/25.04/

    work page doi:10.5194/gmd-13-1335-2020 2020

  43. [43]

    Harris,The ARM ASTC Encoder, a compressor for the Adaptive Scalable Texture Compression data format

    P. Harris,The ARM ASTC Encoder, a compressor for the Adaptive Scalable Texture Compression data format. [Online]. Available: https://github.com/ARM-software/astc-encoder

  44. [44]

    The ASWF Takes Open- ColorIO to the Next Level,

    D. Walker, M. Dolan, and P. Hodoul, “The ASWF Takes Open- ColorIO to the Next Level,” inProceedings of the 2020 Digital Production Symposium, ser. DigiPro ’20, Virtual Event, USA: ACM, 2020,ISBN: 9781450380348.DOI: 10 . 1145 / 3403736 . 3403942 [Online]. Available: www.opencolorio.org

    2020

  45. [45]

    URLhttps://doi.org/10

    C. Geuzaine and J.-F. Remacle, “GMSH: A 3-D finite element mesh generator with built-in pre- and post-processing facilities,”Inter- national Journal for Numerical Methods in Engineering, vol. 79, no. 11, pp. 1309–1331, 2009.DOI: 10.1002/nme.2579 [Online]. Available: www.gmsh.info

    work page doi:10.1002/nme.2579 2009

  46. [46]

    JSBSim: An Open Source Flight Dynamics Model in C++,

    J. Berndt, “JSBSim: An Open Source Flight Dynamics Model in C++,” inAIAA Modeling and Simulation Technologies Conference and Exhibit. AIAA Journal, 2012.DOI: 10 . 2514 / 6 . 2004 - 4923 [Online]. Available: https://github.com/JSBSim-Team/jsbsim

    2012

  47. [47]

    Parallel power computation for photonic crystal devices,

    U. Andersson, M. Qiu, and Z. Zhang, “Parallel power computation for photonic crystal devices,”Methods and Applications of Analysis, vol. 13, pp. 149–156, Jul. 2006.DOI: 10.4310/MAA.2006.v13.n2.a3

    work page doi:10.4310/maa.2006.v13.n2.a3 2006

  48. [48]

    Scalable molecular dynamics with namd,

    J. C. Phillips et al., “Scalable molecular dynamics with namd,”Jour- nal of Computational Chemistry, vol. 26, no. 16, pp. 1781–1802, 2005.DOI: 10.1002/jcc.20289 [Online]. Available: www.ks.uiuc. edu/Research/namd

    work page doi:10.1002/jcc.20289 2005

  49. [49]

    The Regional Oceanic Modeling System (ROMS): a Split-explicit, Free-surface, Topography-following-coordinate Oceanic Model,

    A. F. Shchepetkin and J. C. McWilliams, “The Regional Oceanic Modeling System (ROMS): a Split-explicit, Free-surface, Topography-following-coordinate Oceanic Model,”Ocean Mod- elling, vol. 9, no. 4, 2005,ISSN: 1463-5003.DOI: 10.1016/j.ocemod. 2004.08.002 [Online]. Available: www.myroms.org

    work page doi:10.1016/j.ocemod 2005

  50. [50]

    Kronbichler, D

    M. Kronbichler, D. Sashko, and P. Munch,Enhancing data locality of the conjugate gradient method for high-order matrix-free finite- element implementations, 2022. arXiv: 2205 . 08909[cs.MS]. [Online]. Available: https://github.com/kronbichler/spec-femflow

    2022

  51. [51]

    NEST: The Neural Simulation Tool,

    H. E. Plesser et al., “NEST: The Neural Simulation Tool,” inEncy- clopedia of Computational Neuroscience. New York, NY: Springer New York, 2022, pp. 2187–2189,ISBN: 978-1-0716-1006-0.DOI: 10.1007/978-1-0716-1006-0_258 [Online]. Available: https://nest- initiative.org

    work page doi:10.1007/978-1-0716-1006-0_258 2022

  52. [52]

    Marian: Fast Neural Machine Trans- lation in C++,

    M. Junczys-Dowmunt et al., “Marian: Fast Neural Machine Trans- lation in C++,” 2018. arXiv: 1804.00344. [Online]. Available: https: //marian-nmt.github.io

    work page arXiv 2018

  53. [53]

    Optimization And Profiling Of The Cache Per- formance Of Parallel Lattice Boltzmann Codes,

    T. Pohl et al., “Optimization And Profiling Of The Cache Per- formance Of Parallel Lattice Boltzmann Codes,”Parallel Process- ing Letters, vol. 13, pp. 549–560, Dec. 2003.DOI: 10 . 1142 / S0129626403001501

    2003

  54. [54]

    Performance Analysis of a Reduced Data Movement Algorithm for Neutron Cross Section Data in Monte Carlo Simulations,

    J. R. Tramm et al., “Performance Analysis of a Reduced Data Movement Algorithm for Neutron Cross Section Data in Monte Carlo Simulations,” inSolving Software Challenges for Exascale, Springer International Publishing, 2015, pp. 39–56,ISBN: 978-3- 319-15976-8.DOI: 10 . 1007 / 978 - 3 - 319 - 15976 - 8 _ 3 [Online]. Available: https://github.com/ANL-CESAR

    2015

  55. [55]

    Comparing Performance of C Compilers Optimizations on Different Multicore Architectures,

    R. S. Machado et al., “Comparing Performance of C Compilers Optimizations on Different Multicore Architectures,” in2017 In- ternational Symposium on Computer Architecture and High Perfor- mance Computing Workshops (SBAC-PADW), 2017, pp. 25–30.DOI: 10.1109/SBAC-PADW.2017.13

    work page doi:10.1109/sbac-padw.2017.13 2017

  56. [56]

    GCC vs. ICC comparison using PARSEC Benchmarks,

    A. Almomany, A. Alquraan, and L. Balachandran, “GCC vs. ICC comparison using PARSEC Benchmarks,”International Jour- nal of Innovative Technology and Exploring Engineering, vol. 4, pp. 76–82, Dec. 2014. [Online]. Available: www . researchgate . net / publication / 367255416 _ GCC _ vs _ ICC _ comparison _ using _ PARSEC_Benchmarks

    2014

  57. [57]

    Performance Exploration of Various C/C++ Compilers for AMD EPYC Processors in Numerical Modeling of Solidification,

    K. Halbiniak et al., “Performance Exploration of Various C/C++ Compilers for AMD EPYC Processors in Numerical Modeling of Solidification,”Advances in Engineering Software, vol. 166, p. 103 078, 2022,ISSN: 0965-9978.DOI: 10 . 1016 / j . advengsoft . 2021.103078

    work page arXiv 2022

  58. [58]

    SPEC CPU2017: Performance, Event, and Energy Characterization on the Core i7-8700K,

    R. Hebbar S R and A. Milenkovi ´c, “SPEC CPU2017: Performance, Event, and Energy Characterization on the Core i7-8700K,” in Proceedings of the 2019 ACM/SPEC International Conference on Performance Engineering, ser. ICPE ’19, Mumbai, India: ACM, 2019, 111–118,ISBN: 9781450362399.DOI: 10 . 1145 / 3297663 . 3310314

    2019

  59. [59]

    [On- line]

    SPEC,SPEC CPU ® v8 Benchmark Search Program, 2020. [On- line]. Available: www.spec.org/cpu/cpuv8

    2020

  60. [60]

    SPEC,Step 3 Rules of the CPUv8 Benchmark Search Program,

  61. [61]

    Available: www.spec.org/cpu/cpuv8/#step3

    [Online]. Available: www.spec.org/cpu/cpuv8/#step3

  62. [62]

    [Online]

    ACM,NAMD: First ACM Gordon Bell Special Prize for High Performance Computing-Based COVID-19 Research Awarded, 2020. [Online]. Available: www.acm.org/media- center/2020/november/ gordon-bell-special-prize-covid-research-2020

    2020

  63. [63]

    [Online]

    Community,JSBSim for SPEC CPU v8, JSBSim, 2023. [Online]. Available: https://github.com/JSBSim-Team/jsbsim/issues/834

    2023

  64. [64]

    [Online]

    Community,NEST for SPEC CPU v8, NEST, 2024. [Online]. Available: https://github.com/nest/nest-simulator/issues/3217

    2024

  65. [65]

    [Online]

    Academy of Motion Picture Arts and Sciences,2013 Sci-Tech Awards: Jeremy Selan for OCIO, YouTube, 2014. [Online]. Avail- able: www.youtube.com/watch?v=PMTSbvdVxnE

    2013

  66. [66]

    [Online]

    SPEC,CPU ®2026 Licenses. [Online]. Available: www.spec.org/ cpu2026/Docs/licenses.html

    2026

  67. [67]

    Workload Design: selecting representative program-input pairs,

    L. Eeckhout, H. Vandierendonck, and K. De Bosschere, “Workload Design: selecting representative program-input pairs,” inProceed- ings of International Conference on Parallel Architectures and Compilation Techniques, 2002, pp. 83–94.DOI: 10 . 1109 / PACT. 2002.1106006

    work page arXiv 2002

  68. [68]

    Bartz-Beielstein et al.,Benchmarking in Optimization: Best Practice and Open Issues, 2020

    T. Bartz-Beielstein et al.,Benchmarking in Optimization: Best Practice and Open Issues, 2020. arXiv: 2007.03488[cs.NE]

    work page arXiv 2020

  69. [69]

    The Alberta Workloads for the SPEC CPU 2017 Benchmark Suite,

    J. N. Amaral et al., “The Alberta Workloads for the SPEC CPU 2017 Benchmark Suite,” inIEEE ISPASS, 2018, pp. 159–168.DOI: 14 10.1109/ISPASS.2018.00029 [Online]. Available: https://webdocs. cs.ualberta.ca/~amaral/AlbertaWorkloadsForSPECCPU2017/

    work page doi:10.1109/ispass.2018.00029 2017

  70. [70]

    Evaluating the Evaluations: A Perspec- tive on Benchmarks,

    O. Alonso and K. Church, “Evaluating the Evaluations: A Perspec- tive on Benchmarks,”SIGIR Forum, vol. 58, no. 2, 1–27, Mar. 2025, ISSN: 0163-5840.DOI: 10.1145/3722449.3722467

    work page doi:10.1145/3722449.3722467 2025

  71. [71]

    BetterBench: Assessing AI Benchmarks, Uncover- ing Issues, and Establishing Best Practices,

    A. Reuel et al., “BetterBench: Assessing AI Benchmarks, Uncover- ing Issues, and Establishing Best Practices,” inProceedings of the 38th International Conference on Neural Information Processing Systems, ser. NIPS ’24, Vancouver, BC, Canada: Curran Associates Inc., 2024,ISBN: 9798331314385. arXiv: 2411.12990[cs.AI]

    work page arXiv 2024

  72. [74]

    Wittich,Ampere Strategy and Roadmap Update, 2024

    J. Wittich,Ampere Strategy and Roadmap Update, 2024. [Online]. Available: www.youtube.com/watch?v=cYrT2ohcykk&t=1404s

    2024

  73. [75]

    Brown,AWS re:Invent 2023 - Compute innovation for any application, anywhere, 2023

    D. Brown,AWS re:Invent 2023 - Compute innovation for any application, anywhere, 2023. [Online]. Available: www. youtube . com/watch?v=dxm93_qgRzk&t=2210s

    2023

  74. [76]

    SPEC CPU Suite Growth: an Historical Perspective,

    J. L. Henning, “SPEC CPU Suite Growth: an Historical Perspective,” SIGARCH Comput. Archit. News, vol. 35, no. 1, 65–68, Mar. 2007, ISSN: 0163-5964.DOI: 10.1145/1241601.1241615

    work page doi:10.1145/1241601.1241615 2007

  75. [77]

    In: IEEE European Symp

    R. Panda et al., “Wait of a Decade: Did SPEC CPU 2017 Broaden the Performance Horizon?” In2018 IEEE International Sympo- sium on High Performance Computer Architecture (HPCA), 2018, pp. 271–282.DOI: 10.1109/HPCA.2018.00032

    work page doi:10.1109/hpca.2018.00032 2017

  76. [78]

    DCPerf: An Open-Source, Battle-Tested Performance Benchmark Suite for Datacenter Workloads,

    W. Su et al., “DCPerf: An Open-Source, Battle-Tested Performance Benchmark Suite for Datacenter Workloads,” inProceedings of the 52nd Annual International Symposium on Computer Architec- ture, ser. ISCA ’25, Tokyo, Japan: ACM, 2025, 1717–1730,ISBN: 9798400712616.DOI: 10.1145/3695053.3731411

    work page doi:10.1145/3695053.3731411 2025

  77. [79]

    MisSPECulation: Partial and Misleading Use of SPEC CPU2000 in Computer Architecture Conferences,

    D. Citron, “MisSPECulation: Partial and Misleading Use of SPEC CPU2000 in Computer Architecture Conferences,” inProceedings of the 30th Annual International Symposium on Computer Architecture, San Diego, California: ACM, 2003, 52–61,ISBN: 0769519458.DOI: 10.1145/859618.859625

    work page doi:10.1145/859618.859625 2003

  78. [80]

    The Use and Abuse of SPEC: An ISCA Panel,

    IEEE Micro staff, “The Use and Abuse of SPEC: An ISCA Panel,” IEEE Micro, vol. 23, no. 4, 73–77, Jul. 2003,ISSN: 0272-1732.DOI: 10.1109/MM.2003.1225977

    work page doi:10.1109/mm.2003.1225977 2003

  79. [81]

    Analysis of Redundancy and Application Balance in the SPEC CPU2006 Benchmark Suite,

    A. Phansalkar, A. Joshi, and L. K. John, “Analysis of Redundancy and Application Balance in the SPEC CPU2006 Benchmark Suite,” inProceedings of the 34th Annual International Symposium on Computer Architecture, ser. ISCA ’07, San Diego, California, USA: ACM, 2007, 412–423,ISBN: 9781595937063.DOI: 10 . 1145 / 1250662.1250713

    work page arXiv 2007

  80. [82]

    SPEC as a Performance Evaluation Measure,

    R. Giladi and N. Ahitav, “SPEC as a Performance Evaluation Measure,”Computer, vol. 28, no. 8, pp. 33–42, 1995.DOI: 10 . 1109/2.402073

    1995

Showing first 80 references.