Pith. sign in

REVIEW 3 major objections 4 minor 36 references

ASPO: Constraint-Aware Bayesian Optimization for FPGA-based Soft Processors

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read ASPO claims that Bayesian optimization can be made to respect categorical parameters and hardware constraints by customizing its kernel and acquisition function, yielding faster soft-processor designs in less design time.

desk verdict The system integration is real but the core constraint formula is wrong, so the paper's central claim is unsupported and it should be rejected, with a path to revision. read the letter →

arxiv 2506.06817 v1 pith:JYQJYNBQ submitted 2025-06-07 cs.AR cs.LGcs.NEcs.PF

Haoran Wu , Ce Guo , Wayne Luk , Robert Mullins This is my paper
classification cs.ARcs.LGcs.NEcs.PF
keywords BayesianoptimizationFPGAsoftprocessorscategoricalparametersconstrainthandlingincrementalsynthesisdesignspaceexplorationRISC-V
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

ASPO is an attempt to make Bayesian optimization actually work for FPGA-based soft processors—processor designs written in software and deployed on reconfigurable chips—by changing BO's internals rather than bolting on workarounds. The paper's claim is that three customizations—a covariance kernel that respects categorical parameters such as branch-predictor type, smooth functions that encode if-then and divisibility constraints from the processor specification, and an acquisition function that penalizes candidates far from already-synthesized configurations—let the optimizer find valid, fast designs while spending far less wall-clock time on synthesis. If this holds, designers of soft processors can replace manual tuning with an automated search that respects hardware constraints and reports, on the tested RISC-V processors and seven benchmarks, better designs than the default configuration and competing methods in most tasks. The headline numbers are a 34.6% reduction in execution time on the BOOM processor's multiply benchmark and up to 74% lower total design time than a leading hardware-oriented Bayesian optimization baseline.

What carries the argument

The load-bearing machinery is a modified Gaussian-process Bayesian optimizer with three custom pieces. First, a categorical covariance kernel: one-hot encoded categorical parameters are collapsed to their dominant category before covariance is computed, making similarity depend on actual category identity and preventing redundant sampling of identical designs. Second, smooth constraint-checking functions built from min and max over simple quadratic if-then conditions, plus a sine-based divisibility check, so that conjunctive and disjunctive parameter constraints can guide gradient-based acquisition maximization. Third, a cost-aware acquisition function of the form $\alpha_{\text{cool}}(x,t) = \frac{\alpha(x)}{\lambda(t)\,\hat{c}(x)}$, where $\hat{c}(x)$ is the minimum weighted Euclidean distance from the candidate to a configuration in the synthesis checkpoint database and $\lambda(t)$ decays exponentially over iterations; this is paired with an orthogonal-array warm start and a weighted-distance checkpoint matcher for incremental FPGA synthesis. Together these pieces convert a black-box parameter search into one that respects specification constraints and synthesis cost while exploring the design space.

What would settle it

Measure actual FPGA synthesis times for many pairs of configurations at varying weighted distances and test whether the minimum weighted Euclidean distance to the checkpoint database predicts synthesis time; if the correlation is weak or absent, the cost-aware penalty is steering on a bad proxy and the reported design-time savings would not hold for new processors.

Watch

Extended reading notes

Core claim

The paper's central discovery is that the three obstacles to applying Bayesian optimization to soft processors—categorical parameters, Boolean parameter constraints, and expensive FPGA synthesis—can be absorbed into the BO mechanism itself rather than worked around. Categorical parameters are one-hot encoded and handled by a customized covariance kernel that maps the continuous relaxation back to a hard category assignment, so the optimizer stops resampling identical designs. Constraints are encoded as smooth, differentiable functions using min and max compositions of quadratic if-then conditions and a sine-based divisibility check, so conjunctive and disjunctive specifications become numeric feasibility signals that a gradient-based inner solver can respect. Evaluation cost is folded into a cost-aware acquisition function that penalizes candidates far from already-synthesized configurations, with a decaying penalty schedule, and a checkpoint database plus weighted-distance matcher enables incremental synthesis reuse. The paper reports that this combination yields valid, high-performance configurations faster: a 34.6% execution-time improvement on the BOOM multiply benchmark and design-time reductions up to 74% compared to a state-of-the-art physical-aware BO baseline.

Load-bearing premise

The whole design-time speedup rests on the assumption that the minimum weighted Euclidean distance to an already-synthesized configuration predicts how long FPGA synthesis will take, but the paper presents no measurement of that correlation.

Editorial extensions

If this is right

  • Constraint-heavy processors become tractable: the invalid-design rate drops on most tasks for the larger processors, so designers waste less time on configurations that fail specification or resource checks.
  • Design time scales better with processor complexity: the savings are largest on the processor with the longest per-evaluation synthesis time, where total design time falls by up to 74% relative to a leading physical-aware BO approach.
  • Design quality does not suffer for the speedup: ASPO-optimized configurations had the best estimated execution time in 17 of 21 processor-benchmark tasks, including a 34.6% improvement on the multiply benchmark.
  • Because the framework exposes a configuration interface and supports designer-specified processors, the same constraint encoding and checkpoint reuse can be applied to other configurable RISC-V soft cores with modest modification.

Reading between the lines

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

  • The weighted-distance synthesis-time proxy is never validated against measured synthesis times; if the proxy is poor, the cost-aware penalty could bias the search toward cheap-to-synthesize but mediocre configurations, and a direct correlation study would settle this.
  • The customized categorical kernel and min/max constraint encoding are generic mechanisms that could transfer to other constrained hardware tuning problems, such as high-level synthesis pragma search or accelerator configuration, where the same categorical-plus-constraint structure appears.
  • The cooling schedule's two parameters are set empirically, so the advantage may be sensitive to them; an adaptive schedule driven by observed synthesis times might remove that tuning burden.
  • The reported design-time savings conflate two effects—checkpoint reuse and the cost-aware acquisition penalty—so an ablation separating them would show which one drives the improvement.
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 / 4 minor

Summary. The paper presents ASPO, a Bayesian optimization (BO) framework for tuning FPGA-based soft processors. ASPO introduces three modifications to standard BO: a covariance kernel customized for categorical parameters, smooth constraint functions to encode parameter dependencies, and a cost-aware acquisition function that penalizes candidate configurations that are far from previously synthesized checkpoints, enabling reuse of incremental synthesis results. The approach is evaluated on three RISC-V soft processors (EL2 VeeR, RocketChip, BOOM) using seven benchmarks, reporting reductions in estimated execution time (EET) and total design time (TDT) compared to baselines such as VBO, BOOM-Explorer, RCBO, and Boomerang.

Significance. If the claims hold, ASPO would be a practically useful contribution to FPGA design-space exploration, particularly for combining categorical parameters, parameter constraints, and synthesis-cost awareness in a BO framework. The paper provides open-source artifacts and targets real soft-processor designs, which increases its potential impact. However, the current manuscript contains a clear mathematical error in the core conditional-constraint encoding, which undermines the central 'constraint-aware' claim. The reported quantitative results are therefore not grounded in the described algorithm, and the checkpoint-proxy mechanism is not empirically validated. The paper's contribution is potentially salvageable, but a major revision with corrected formulations and re-run experiments is needed before the claims can be assessed.

major comments (3)
  1. [Section III-B, Eq. (2)-(3)] The constraint function C_b(x) = min_m c_m(x_m) with c_m(x_m) = -(x_m - a_m)(x_m - b_m) does not implement the conditional constraint 'if x1 in [a1,b1] then x2 in [a2,b2]'. Since c_m is nonnegative exactly when x_m lies in its interval, C_b is nonnegative if and only if both x1 and x2 lie in their respective intervals. This is a conjunction, not an implication. A configuration with x1 outside [a1,b1] has c1(x1) < 0, hence C_b < 0, and is rejected even though the conditional constraint is vacuously satisfied. A correct smooth encoding would be C_b(x) = max(-c_1(x_1), c_2(x_2)), which is nonnegative when x1 is outside the antecedent interval or when x2 is inside the consequent interval. This error propagates through the disjunctive and conjunctive combinators in Eq. (4)-(5), and therefore all constraints listed in Table III are mis-encoded by the described method. Because the feasibility function is central to the algorithm, the reported Invalid Design Rate (IDR), Total Design Time (TDT), and Estimated Execution Time (EET) results in Tables VI and VII are obtained with a different, more restrictive feasible region than the one specified. The paper's central claim of constraint-aware optimization is unsupported as written.
  2. [Section IV-A, Eq. (7)-(10)] The cost-aware acquisition function and the checkpoint configuration matcher assume that the weighted Euclidean distance d(x,q) is a reliable proxy for FPGA synthesis time. However, the paper provides no experiment or measurement showing this correlation. The weights w* are optimized on a small random subset (Eq. 8) and the cooling parameters lambda0 and k are 'determined empirically through experiments', but no sensitivity or validation is reported. Without such evidence, the reported TDT savings (e.g., up to 74% vs. Boomerang) could be dominated by the growth of the checkpoint database or by changes in the search distribution rather than by the cost-aware mechanism. This is a load-bearing point for the design-time improvement claims and requires either a direct correlation study or ablation experiments demonstrating the mechanism's effect.
  3. [Section III-A] The description of the categorical covariance kernel is internally unclear and appears technically incorrect as stated. The text says that after one-hot encoding and normalization, 'the covariance function sigma^2(x,x'') for any sample that can be transformed to the same one-hot vector becomes zero' and that this 'prevents the repetitive selection of identical soft processor designs'. If the covariance between an evaluated point and a candidate that maps to the same one-hot vector is zero, the Gaussian process treats those points as uncorrelated, which would not discourage re-selection; it would instead treat the candidate as unexplored. A kernel value of one (or a deterministic tie) would be needed to achieve the claimed behavior. The authors should clarify the intended kernel definition and correct the description; as written, this part of the algorithm does not support the stated objective.
minor comments (4)
  1. [Section I, Introduction] The acronym 'ASPO' is written as 'ASOP' in the fourth paragraph; please correct the typo.
  2. [Table VI] The dagger symbol (†) for Hill Climbing TDT values is not explained in the table caption or the surrounding text; clarify whether HC's TDT is excluded from comparison or reported for completeness.
  3. [Section IV-A, Eq. (8)] The optimization of w* is stated to use 'heuristic search techniques' on a random subset, but no details of the heuristic, its convergence, or the subset size are given; this makes the checkpoint-matching results difficult to reproduce.
  4. [Section V-C, Table VI] The paper reports IDR and TDT for each method, but the experimental setup does not state the number of BO iterations or the stopping criterion used by each method; please specify these to allow interpretation of the convergence times.

Circularity Check

0 steps flagged · score 2.0 of 10

No load-bearing circularity; the paper's main performance and design-time claims are measured against external baselines and not built into the objective. The only self-citation is minor and non-load-bearing.

full rationale

ASPO's central claims (34.6% EET reduction on BOOM multiply, up to 74% TDT reduction versus Boomerang) are empirical comparisons: EET is computed from simulated cycles and synthesis-timing reports, and TDT is measured wall-clock time for the whole design loop. Neither quantity appears as a fitted constant in the acquisition function or surrogate. The cost-aware penalty uses the checkpoint-distance proxy of Eq. (7), with weights w* optimized on a random subset in Eq. (8) and lambda0/k chosen empirically; these are ordinary hyperparameters and not the reported results, so this is not a fitted input renamed as a prediction. The categorical-kernel mechanism in Section III-A cites the authors' prior RCBO work [10] for the zero-covariance property of identical one-hot-transformed vectors, but the property follows immediately from the transformation described in the same section, so the self-citation is not load-bearing. The strongest in-text caveat is a correctness issue rather than circularity: C_b(x)=min_m c_m(x_m) in Eq. (2)-(3) encodes a conjunction ("x1 in interval AND x2 in interval") instead of the intended implication ("if x1 in interval THEN x2 in interval"), so the constraint-awareness claim is not fully supported as written. That flaw does not make any output equal to an input by construction, so it does not raise the circularity score, though it is a substantive correctness risk.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The framework relies on standard BO/GP assumptions, an unvalidated distance-to-synthesis-time proxy, and a logically flawed constraint formula. The free parameters are engineering hyperparameters fitted to data rather than physical constants, and no new theoretical entities are introduced.

free parameters (3)
  • lambda0 (initial cost sensitivity) = not reported
    In the cooling schedule Eq. (10), lambda0 and k are tuned empirically through experiments to balance exploration and evaluation cost; the chosen values are not disclosed.
  • k (cost decay rate) = not reported
    Same as lambda0; determines how quickly the penalty on distant configurations decays during optimization.
  • w (checkpoint matcher weights) = not reported
    The weight vector in Eq. (7) is optimized via Eq. (8) to minimize synthesis time on a randomly sampled subset, making the distance metric specifically fitted to synthesis-time data.
assumptions (5)
  • standard math Standard Gaussian process and Bayesian optimization machinery are valid for this design space.
    The paper assumes GP surrogates and Expected Improvement acquisition work as usual on the transformed continuous space, without re-deriving them.
  • domain assumption The smooth penalty functions defined in Section III-B can be embedded in a gradient-based acquisition optimizer to enforce constraints.
    The paper assumes that maximizing the acquisition subject to PC(x) >= 0 yields feasible configurations, yet the implication encoding in Eq. (2)-(3) is logically incorrect.
  • domain assumption Incremental synthesis using stored checkpoints produces correct bitstreams and valid timing reports.
    The acceleration method relies on the FPGA tool's incremental synthesis behavior, citing [20], and assumes reusing a near-match checkpoint does not degrade result correctness.
  • domain assumption Weighted Euclidean distance in parameter space is a suitable proxy for synthesis-time similarity.
    Equation (7) is used both for checkpoint selection and in the cost-aware acquisition function, but no direct evidence links this distance to actual synthesis runtime.
  • standard math Orthogonal array sampling provides uniform coverage of the design space.
    The warm-start mechanism assumes OA-generated samples evenly cover the space, citing [21], without validation on these specific design spaces.

how reviews work

0 comments
Cite this review

Pith. "Pith review of ASPO: Constraint-Aware Bayesian Optimization for FPGA-based Soft Processors." pith.science (2026). https://pith.science/paper/JYQJYNBQ

@misc{pith2026250606817,
  author       = {Pith},
  title        = {Pith review of: ASPO: Constraint-Aware Bayesian Optimization for FPGA-based Soft Processors},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JYQJYNBQ}},
  note         = {Machine review of arXiv:2506.06817}
}
read the original abstract

Bayesian Optimization (BO) has shown promise in tuning processor design parameters. However, standard BO does not support constraints involving categorical parameters such as types of branch predictors and division circuits. In addition, optimization time of BO grows with processor complexity, which becomes increasingly significant especially for FPGA-based soft processors. This paper introduces ASPO, an approach that leverages disjunctive form to enable BO to handle constraints involving categorical parameters. Unlike existing methods that directly apply standard BO, the proposed ASPO method, for the first time, customizes the mathematical mechanism of BO to address challenges faced by soft-processor designs on FPGAs. Specifically, ASPO supports categorical parameters using a novel customized BO covariance kernel. It also accelerates the design evaluation procedure by penalizing the BO acquisition function with potential evaluation time and by reusing FPGA synthesis checkpoints from previously evaluated configurations. ASPO targets three soft processors: RocketChip, BOOM, and EL2 VeeR. The approach is evaluated based on seven RISC-V benchmarks. Results show that ASPO can reduce execution time for the ``multiply'' benchmark on the BOOM processor by up to 35\% compared to the default configuration. Furthermore, it reduces design time for the BOOM processor by up to 74\% compared to Boomerang, a state-of-the-art hardware-oriented BO approach.

Figures

Figures reproduced from arXiv: 2506.06817 by the authors.

Figure 1
Figure 1. Besides processor optimization, BO has been applied be￾yond architectural parameter tuning to various aspects of hardware and software optimization on FPGA platforms. For instance, BO has been employed to refine computational modules, such as in [11], where it is used to generate approximate multipliers that balance accuracy and hardware efficiency. Similarly, [12] demonstrates the application of BO to determine opt… view at source ↗
Figure 1
Figure 1. This comparison evaluates the workflows of ASPO, RCBO [10], BOOM-Explorer [3], and vanilla BO [9]. Compared to RCBO, ASPO incorporates an [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 3
Figure 3. Accelerated evaluation flow. To take advantage of the incremental synthesis feature for faster evaluation, two key components are required. The first is a database of checkpoints to store designs that can serve as starting points for incrementally synthesizing new configurations. The second is a matching function to identify an appropriate starting point from the database, minimizing potential synthesis time by reus… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: The two frameworks employing acceleration during the logic synthesis [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

36 extracted references · 30 canonical work pages

  1. [1]

    Soft-core processors for embedded systems,

    J. G. Tong, I. D. L. Anderson, and M. A. S. Khalid, “Soft-core processors for embedded systems,”2006 International Conference on Microelectronics, pp. 170–173, 2006. [Online]. Available: https://api.semanticscholar.org/CorpusID:5876189

    work page 2006

  2. [2]

    Dealing with categorical and integer-valued variables in Bayesian Optimization with Gaussian processes,

    E. C. Garrido-Merch ´an and D. Hern ´andez-Lobato, “Dealing with categorical and integer-valued variables in Bayesian Optimization with Gaussian processes,”Neurocomputing, vol. 380, p. 20–35, Mar. 2020. [Online]. Available: http://dx.doi.org/10.1016/j.neucom.2019.11.004

    doi:10.1016/j.neucom.2019.11.004 2020

  3. [3]

    BOOM-Explorer: RISC-V BOOM Microarchitecture Design Space Exploration,

    C. Bai, Q. Sun, J. Zhai, Y . Ma, B. Yu, and M. D. F. Wong, “BOOM-Explorer: RISC-V BOOM Microarchitecture Design Space Exploration,”ACM Trans. Des. Autom. Electron. Syst., vol. 29, no. 1, dec 2023. [Online]. Available: https://doi.org/10.1145/3630013

    doi:10.1145/3630013 2023

  4. [4]

    Boomerang: Physical-Aware De- sign Space Exploration Framework on RISC-V SonicBOOM Microar- chitecture,

    Y .-F. Liu, C.-Y . Hsieh, and S.-Y . Kuo, “Boomerang: Physical-Aware De- sign Space Exploration Framework on RISC-V SonicBOOM Microar- chitecture,” in2023 IEEE 34th International Conference on Application- specific Systems, Architectures and Processors (ASAP), 2023, pp. 85–93

    work page 2023

  5. [5]

    Taking the human out of the loop: A review of Bayesian optimization,

    B. Shahriari, K. Swersky, Z. Wang, R. P. Adams, and N. De Freitas, “Taking the human out of the loop: A review of Bayesian optimization,” Proceedings of the IEEE, vol. 104, no. 1, pp. 148–175, 2015

    work page 2015

  6. [6]

    C. K. Williams and C. E. Rasmussen,Gaussian processes for machine learning. MIT press Cambridge, MA, 2006, vol. 2, no. 3

    2006

  7. [7]

    FIST: A feature-importance sampling and tree-based method for automatic design flow parameter tuning,

    Z. Xie, G.-Q. Fang, Y .-H. Huang, H. Ren, Y . Zhang, B. Khailany, S.-Y . Fang, J. Hu, Y . Chen, and E. C. Barboza, “FIST: A feature-importance sampling and tree-based method for automatic design flow parameter tuning,” in2020 25th Asia and South Pacific Design Automation Conference (ASP-DAC). IEEE, 2020, pp. 19–25

    work page 2020

  8. [8]

    Exploring machine learning adoption in customisable processor design,

    J. G. Coutinho, C. Guo, T. Todman, and W. Luk, “Exploring machine learning adoption in customisable processor design,” inInternational Conference on ASIC (ASICON). IEEE, 2023, pp. 1–4

    work page 2023

Show all 36 references
  1. [9]

    Is Vanilla Bayesian Optimization Enough for High-Dimensional Architec- ture Design Optimization?

    Y . Gao, D. Luo, C. Bai, B. Yu, H. Geng, Q. Sun, and C. Zhuo, “Is Vanilla Bayesian Optimization Enough for High-Dimensional Architec- ture Design Optimization?” inICCAD, 2024

    2024

  2. [10]

    Resource-Constraint Bayesian Optimiza- tion for Soft Processors on FPGAs,

    C. Guo, H. Wu, and W. Luk, “Resource-Constraint Bayesian Optimiza- tion for Soft Processors on FPGAs,” in14th International Symposium on Highly Efficient Accelerators and Reconfigurable Technologies (HEART ’24). Porto, Portugal: ACM, June 19-21 2024, pp. 1–13

    2024

  3. [11]

    AMG: Automated Efficient Ap- proximate Multiplier Generator for FPGAs via Bayesian Optimization,

    Z. Li, H. Zhou, L. Wang, and X. Zhou, “AMG: Automated Efficient Ap- proximate Multiplier Generator for FPGAs via Bayesian Optimization,” in2023 International Conference on Field Programmable Technology (ICFPT), 2023, pp. 294–295

    2023

  4. [12]

    Layer-Specific Optimization for Mixed Data Flow With Mixed Precision in FPGA Design for CNN- Based Object Detectors,

    D. T. Nguyen, H. Kim, and H.-J. Lee, “Layer-Specific Optimization for Mixed Data Flow With Mixed Precision in FPGA Design for CNN- Based Object Detectors,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 31, no. 6, pp. 2450–2464, 2021

    2021

  5. [13]

    Graebo: Fpga general routing architecture exploration via Bayesian optimization,

    S. Zheng, J. Qian, H. Zhou, and L. Wang, “Graebo: Fpga general routing architecture exploration via Bayesian optimization,” in2022 32nd In- ternational Conference on Field-Programmable Logic and Applications (FPL). IEEE, 2022, pp. 282–286

    2022

  6. [14]

    Multi-objective design space exploration for high-level synthesis via Bayesian optimization,

    H. Kuang and L. Wang, “Multi-objective design space exploration for high-level synthesis via Bayesian optimization,” in2023 International Symposium of Electronics Design Automation (ISEDA). IEEE, 2023, pp. 150–155

    2023

  7. [15]

    Hgbo-dse: Hierarchical gnn and Bayesian optimization based hls design space exploration,

    H. Kuang, X. Cao, J. Li, and L. Wang, “Hgbo-dse: Hierarchical gnn and Bayesian optimization based hls design space exploration,” in2023 International Conference on Field Programmable Technology (ICFPT). IEEE, 2023, pp. 106–114

    2023

  8. [16]

    A Tutorial on Bayesian Optimization of Expensive Cost Functions, with Application to Active User Modeling and Hierarchical Reinforcement Learning,

    E. Brochu, V . M. Cora, and N. de Freitas, “A Tutorial on Bayesian Optimization of Expensive Cost Functions, with Application to Active User Modeling and Hierarchical Reinforcement Learning,”CoRR, vol. abs/1012.2599, 2010. [Online]. Available: http://arxiv.org/abs/1012.2599

    2010 arXiv

  9. [17]

    A software package for sequential quadratic programming,

    D. Kraft, “A software package for sequential quadratic programming,” DLR German Aerospace Center – Institute for Flight Mechanics, K ¨oln, Germany, Tech. Rep. DFVLR-FB 88-28, 1988

    1988

  10. [18]

    An Interior Point Algorithm for Large-Scale Nonlinear Programming,

    R. H. Byrd, M. E. Hribar, and J. Nocedal, “An Interior Point Algorithm for Large-Scale Nonlinear Programming,”SIAM J. Optim., vol. 9, pp. 877–900, 1999. [Online]. Available: https://api.semanticscholar.org/CorpusID:16293345

    1999

  11. [19]

    Taking the Human Out of the Loop: A Review of Bayesian Optimiza- tion,

    B. Shahriari, K. Swersky, Z. Wang, R. P. Adams, and N. de Freitas, “Taking the Human Out of the Loop: A Review of Bayesian Optimiza- tion,”Proceedings of the IEEE, vol. 104, no. 1, pp. 148–175, 2016

    2016

  12. [20]

    Towards agile DNN accelerator design using incremental synthesis on fpgas,

    Q. Xiao and Y . Liang, “Towards agile DNN accelerator design using incremental synthesis on fpgas,” inProceedings of the 2022 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays, ser. FPGA ’22. New York, NY , USA: Association for Computing Machinery, 2022, p. 4...

    2022

  13. [21]

    Efficient design space exploration via statistical sampling and AdaBoost learning,

    D. Li, S. Yao, Y .-H. Liu, S. Wang, and X.-H. Sun, “Efficient design space exploration via statistical sampling and AdaBoost learning,” in 2016 53nd ACM/EDAC/IEEE Design Automation Conference (DAC), 2016, pp. 1–6

    2016

  14. [22]

    Cost-aware Bayesian optimization,

    E. H. Lee, V . Perrone, C. Archambeau, and M. Seeger, “Cost-aware Bayesian optimization,” 2020. [Online]. Available: https://arxiv.org/abs/2003.10870

    2020 arXiv

  15. [23]

    Exploring coremark a benchmark maximizing simplicity and efficacy,

    S. Gal-On and M. Levy, “Exploring coremark a benchmark maximizing simplicity and efficacy,”The Embedded Microprocessor Benchmark Consortium, 2012

    2012

  16. [24]

    Dhrystone: a synthetic systems programming benchmark,

    R. P. Weicker, “Dhrystone: a synthetic systems programming benchmark,”Commun. ACM, vol. 27, no. 10, p. 1013–1030, Oct. 1984. [Online]. Available: https://doi.org/10.1145/358274.358283

    1984

  17. [25]

    riscv-software-src/riscv-tests risc-v,

    T. Newsome, “riscv-software-src/riscv-tests risc-v,” https://github.com/riscv-software-src/riscv-tests, May 2019, [Online; accessed January 17, 2025]

    2019

  18. [26]

    The Rocket Chip Generator,

    K. Asanovi ´c, R. Avi ˇzienis, J. Bachrach, S. Beamer, D. Biancolin, C. Celio, H. Cook, P. Dabbelt, J. Hauser, A. Izraelevitz, S. Karandikar, B. Keller, D. Kim, J. Koenig, Y . Lee, E. Love, M. Maas, A. Magyar, H. Mao, M. Moreto, A. Ou, D. Patterson, B. Richards, C. Schmidt, S....

    2016

  19. [27]

    The Berkeley Out-of-Order Machine (BOOM): An Industry-Competitive, Synthesizable, Parameter- ized RISC-V Processor,

    C. Celio, D. A. Patterson, and K. Asanovi ´c, “The Berkeley Out-of-Order Machine (BOOM): An Industry-Competitive, Synthesizable, Parameter- ized RISC-V Processor,” EECS Department, University of California, Berkeley, Tech. Rep. UCB/EECS-2015-167, Jun 2015. [Online]. Available:...

    2015

  20. [28]

    Cores-VeeR-EL2: A RISC-V Core,

    Chips Alliance, “Cores-VeeR-EL2: A RISC-V Core,” https://github.com/chipsalliance/Cores-VeeR-EL2, 2024, available online: https://github.com/chipsalliance/Cores-VeeR-EL2

    2024

  21. [29]

    Ibex: An embedded 32-bit risc-v cpu core,

    lowRISCC.I.C., “Ibex: An embedded 32-bit risc-v cpu core,” https://ibex-core.readthedocs.io, 2021, documentation

    2021

  22. [30]

    The Cost of Application-Class Processing: Energy and Performance Analysis of a Linux-Ready 1.7-GHz 64-Bit RISC-V Core in 22-nm FDSOI Technology,

    F. Zaruba and L. Benini, “The Cost of Application-Class Processing: Energy and Performance Analysis of a Linux-Ready 1.7-GHz 64-Bit RISC-V Core in 22-nm FDSOI Technology,”IEEE Transactions on Very Large Scale Integration (VLSI) Systems, vol. 27, no. 11, pp. 2629–2640, 2019

    2019

  23. [31]

    Can hardware performance counters be trusted?

    V . M. Weaver and S. A. McKee, “Can hardware performance counters be trusted?” in2008 IEEE International Symposium on Workload Characterization, 2008, pp. 141–150

    2008

  24. [32]

    Automating Optimiza- tion of Reconfigurable Designs,

    M. Kurek, T. Becker, T. C. P. Chau, and W. Luk, “Automating Optimiza- tion of Reconfigurable Designs,” inProceedings of the 2014 IEEE 22nd International Symposium on Field-Programmable Custom Computing Machines, ser. FCCM ’14. USA: IEEE Computer Society, 2014, p. 210–213

    2014

  25. [33]

    Gamma: Automating the HW mapping of DNN models on accelerators via genetic algorithm,

    S.-C. Kao and T. Krishna, “Gamma: Automating the HW mapping of DNN models on accelerators via genetic algorithm,” inInternational Conference on Computer-Aided Design, 2020, pp. 1–9

    2020

  26. [34]

    Extended ant colony optimization algorithm for power electronic circuit design,

    J. Zhang, H. S.-H. Chung, A. W.-L. Lo, and T. Huang, “Extended ant colony optimization algorithm for power electronic circuit design,”IEEE Transactions on Power Electronics, vol. 24, no. 1, pp. 147–162, 2008

    2008

  27. [35]

    Towards automated risc- v microarchitecture design with reinforcement learning,

    C. Bai, J. Zhai, Y . Ma, B. Yu, and M. D. Wong, “Towards automated risc- v microarchitecture design with reinforcement learning,” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 38, no. 1, 2024, pp. 12–20

    2024

  28. [36]

    Selection of the Register File Size and the Resource Allocation Policy on SMT Processors,

    J. Alastruey, T. Monreal, F. Cazorla, V . Vi˜nals, and M. Valero, “Selection of the Register File Size and the Resource Allocation Policy on SMT Processors,” in2008 20th International Symposium on Computer Archi- tecture and High Performance Computing. Washington, DC, United S...

    2008

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.