Pith. sign in

REVIEW 3 major objections 5 minor 89 references

Design and accuracy trade-offs in Computational Statistics

T0 review · 3 major / 5 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read This paper establishes that posit arithmetic dominates log-space for statistical computations on extremely small probabilities, improving accuracy, resource use, and speed on two bioinformatics accelerators.

desk verdict Serious empirical look at posit vs log-space for statistical computation; accuracy claim is real on the paper's metric but downstream decisions are never tested. read the letter →

arxiv 2509.10934 v1 pith:LBFX44AG submitted 2025-09-13 math.NA cs.ARcs.NA

Tiancheng Xu , Alan L. Cox , Scott Rixner This is my paper · ORCID
classification math.NAcs.ARcs.NA MSC 65G5065Y04
keywords positarithmeticlog-spacecomputationnumericalunderflowstatisticalcomputinghiddenMarkovmodelPoissonbinomialdistributionFPGAacceleratorfloating-pointaccuracy
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

Statistical computations often multiply probabilities until they become far smaller than ordinary double precision can represent, so software shifts to log-space, trading away arithmetic simplicity. This paper argues that posit, a number format whose exponent and fraction fields expand and contract on demand, avoids that trade-off: it represents the same tiny probabilities directly while keeping more bits for precision. The authors build FPGA accelerators for two bioinformatics workloads of this kind and report that the posit versions are up to two orders of magnitude more accurate, use up to 60% fewer hardware resources, and run up to 33% faster than log-space versions. The central claim is that future hardware for statistical computing should use posits rather than log-space.

What carries the argument

The mechanism is the posit encoding. A posit(N,ES) number reserves one sign bit and then allocates the remaining bits among a run-length-encoded regime, an exponent of ES bits, and a fraction field, so that tiny values automatically receive more regime/exponent bits and values near one receive more fraction bits. Its value is (-1)^sign * (2^(2^ES))^k * 2^e * (1+f). This dynamic allocation is what lets a single 64-bit format cover exponents down to roughly -2,000,000 or beyond while preserving up to 52 fraction bits in the normal range. In the paper's designs, replacing log-sum-exp additions with plain posit multiply-add operations removes the expensive logarithm/exponential units and shorten

What would settle it

Re-run the two application accelerators on the same genomic data using both posit and log-space arithmetic and compare the actual variant calls or evolutionary parameter estimates produced by each; if the calls and estimates agree on essentially all inputs, then the claimed two-orders-of-magnitude accuracy improvement has no observable effect on the applications' outputs. A weaker falsifier: compute the maximum absolute error of the log-likelihood/p-value in log space; if it is comparable or better for log-space, the linear relative-error metric was the source of the apparent advantage.

Watch

Extended reading notes

Core claim

Using logarithms in statistical code is so standard that its numerical cost goes unnoticed: inside the range where binary64 works, log-space is less accurate than binary64 itself, because the exponent field of a log value is mostly zeros while the fraction field has to encode the original exponent. The paper's substantive claim is that this cost is avoidable. Posit—a 64-bit format with run-length-coded scale bits—covers probabilities as small as about 2^-2,000,000 without underflow and still keeps up to 52 fraction bits for ordinary values. On FPGA accelerators for two iterative probability computations, replacing log-space with posit gives final relative errors about 100x smaller, uses roug

Load-bearing premise

The whole comparison rests on measuring accuracy as relative error in the original probability space; if the applications' real outputs—likelihood ratios, p-value thresholds, or parameter estimates—should instead be judged in log space or by the decisions they drive, the reported posit accuracy advantage may not carry through.

Editorial extensions

If this is right

  • Applications that currently compute hidden Markov model likelihoods or Poisson-binomial p-values in log-space can be reimplemented directly in posit arithmetic without losing the tiny values.
  • A posit-based accelerator for these workloads fits roughly 2.5x more processing units on the same FPGA die, because it uses about 40% of the LUTs of a log-space unit.
  • At the same 300 MHz clock, posit units are 15-33% faster, so the throughput per unit of hardware resource about doubles.
  • Final likelihoods and p-values come out with about two orders of magnitude smaller relative error, for example nearly all VICAR-style likelihoods under 1e-8 relative error versus a few percent for log-space.
  • The exponent-size parameter ES can be tuned: larger ES extends range for extremely small values but reduces precision for values near one, so configuration is a real design knob.

Reading between the lines

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

  • A testable extension the paper leaves implicit: evaluate whether the linear-space accuracy gain changes actual variant calls or parameter estimates, since those decisions depend on log-space thresholds.
  • The arithmetic-level analysis is application-independent, so the same design argument likely carries over to other underflow-prone domains such as financial option pricing or probabilistic machine learning; that is an extrapolation, not a claim in the paper.
  • The hardware comparison is on FPGAs at 64-bit width; extending to 32/16-bit formats or ASIC implementations could change the relative resource and latency numbers.
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 / 5 minor

Summary. The paper argues that the standard log-space approach to statistical computations on extremely small probabilities is suboptimal in numerical accuracy, performance, and hardware cost, and that the posit format is a better strategy. It compares binary64, log-space arithmetic, and three posit configurations at the level of individual arithmetic operations, then builds FPGA accelerators for two bioinformatics applications (VICAR's HMM forward algorithm and LoFreq's Poisson-binomial p-value computation). The reported results claim up to two orders of magnitude higher accuracy, up to 60% lower resource use, up to 33% higher performance, and about 2x performance per resource unit versus log-space accelerators. The central accuracy claim is based on relative error of final likelihoods/p-values against 256-bit MPFR results.

Significance. If fully substantiated, the paper would be a useful contribution to numerical analysis and FPGA acceleration for underflow-prone statistical workloads. It provides a quantitative comparison of three number representations, builds optimized accelerators for two real applications, and shows a plausible mechanism by which posit's dynamic exponent/fraction allocation could improve accuracy over log-space. The paper also gives credit to careful use of real datasets (SARS-CoV-2, HCG) and an MPFR ground truth, which strengthens the empirical component. However, the practical significance of the accuracy advantage is not yet demonstrated: the evaluation stops at relative error in linear probability space and never checks whether downstream variant calls or evolutionary parameter estimates change. The resource comparison is also partially confounded by different implementation flows. These gaps are fixable within the manuscript's scope, but they are load-bearing for the broad claim that posit is a 'better strategy' for statistical computations.

major comments (3)
  1. [VI-A and VI-D] The accuracy evaluation is entirely in terms of relative error of final likelihoods/p-values against 256-bit MPFR. LoFreq's output is a binary decision (variant if p < 2^-200), and VICAR's output is evolutionary parameter estimates from variational inference; neither downstream output is examined. Relative error in linear probability is approximately absolute error in log-likelihood, so the metric is not unreasonable, but it does not establish that the reported two-orders-of-magnitude accuracy advantage changes any call or parameter estimate. Please add a call-level comparison for LoFreq (e.g., true/false positives/negatives relative to MPFR) and a comparison of estimated evolutionary parameters for VICAR, for both log-space and posit accelerators. Without this, the conclusion in Section VIII that posit 'leads to improvements in all metrics' overstates what is actually measured.
  2. [VI-D, para. 'However, posits do not always...'] The text discloses that posit(64,12) underflows on 2 of 222,131 p-values and has relative errors up to 10^2129 on 2 others, and posit(64,9) underflows on 132 p-values. Since LoFreq's decision rule is a threshold, these outlier cases could change variant calls. The manuscript does not report whether these p-values lie in the critical set (p < 2^-200) or whether the resulting calls agree with the MPFR baseline. For example, if underflow forces p = 0, the call is always 'variant'; if the true p-value is also below the threshold this is harmless, but if not it is a false positive. A call-level analysis of these extreme cases is necessary to bound the practical impact of the reported accuracy advantage.
  3. [IV-B, Table II] The resource comparison between log-space and posit arithmetic units is confounded by implementation style: the log-space LSE unit uses Xilinx LogiCORE IP (optimized RTL), while the posit units use the MArTo HLS library. The paper acknowledges this difference, but the application-level resource claims in Tables III and IV inherit the same confound, since the log-based accelerators use the Xilinx IP and the posit-based accelerators use MArTo. Because lower resource use is a central part of the claimed advantage, please provide at least one point of comparison with a common design flow (e.g., both in HLS, or a hand-optimized RTL posit implementation), or a sensitivity analysis showing the resource conclusions persist.
minor comments (5)
  1. [IV-A] The section heading contains a typo: 'Numercial Accuracy' should be 'Numerical Accuracy'.
  2. [Figure 3] The y-axis starts at 10^8, so relative errors below 10^-8 are not visible in the box plots. Consider indicating the plot range or adding a zoomed panel, since many application-level errors are in this range.
  3. [Figure 11(b)] The label 'less important' for p-values >= 2^-200 is vague. Since these are non-critical columns, the relevant question is whether any are incorrectly called as variants; please define the label in the caption or text.
  4. [Equation (4)] The definition of l (the number of repeated regime bits) is given verbally but not formally. A compact definition, e.g., in terms of the run-length of the leading regime bit, would improve precision.
  5. [V-A] The authors state that 40% and 5% of critical columns have p-value < 2^-1074 and 2^-10000, respectively, but no distribution is given for p-values close to the decision threshold 2^-200. Such information would help interpret the outlier analysis requested in the major comments.

Circularity Check

1 steps flagged · score 2.0 of 10

Minor tautological config-selection step; central empirical comparison is self-contained.

  1. fitted input called prediction [Section III 'Posit Configuration'; Section VI-D 'Application Numerical Accuracy']
    "Posit(64,18) was selected because it has a sufficient range for extremely small numbers as observed in critical statistical bioinformatics applications. ... Posit(64, 18) does not underflow."

    The ES parameter of posit(64,18) is chosen precisely because its dynamic range (smallest positive about 2^{-16,252,928}, Table I) covers the smallest numbers observed in the target applications (e.g., 2^{-2,900,000} for VICAR and 2^{-434,916} for LoFreq). Therefore the later observation that posit(64,18) 'does not underflow' is a logical consequence of the configuration choice, not an independent empirical result. The central 'two orders of magnitude higher accuracy' claim is not fully circular, since relative error also depends on fraction bits and rounding behavior, but this particular no-underflow supporting observation is forced by construction.

full rationale

The paper is an empirical comparison, not a derivation: arithmetic-level accuracy is measured against 256-bit MPFR ground truth, and application-level accuracy is measured against MPFR-computed likelihoods and p-values. No equation reduces the claimed accuracy benefit to a fitted constant, and the posit configurations are disclosed and varied rather than hidden. The one caveat is the choice of posit(64,18), whose ES is selected to cover the observed range of tiny probabilities; its failure to underflow on those inputs is therefore entailed by the design choice rather than discovered by evaluation. This is a minor, transparent tautology, not a load-bearing circularity. The log-space baseline accelerator comes from the authors' prior work [86], but it is used as a measured benchmark, not as an unverified premise, so this self-citation does not make the argument circular. The use of linear-space relative error as the accuracy metric is a modeling decision and could be debated for decision-focused applications, but it does not assume the paper's conclusion. Overall, the central performance, resource, and accuracy claims retain independent empirical content.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central results rest on three kinds of premises: the definition and behavior of the number formats (posit, binary64, log-space), the choice of MPFR as ground truth, and the application-specific assumption that linear relative error is the right metric. No parameters are fitted to data; the posit ES values are design choices.

free parameters (2)
  • posit ES configurations = 9, 12, 18
    Chosen to represent a range/precision tradeoff, matched to application dynamic ranges. Not fitted to accuracy results, but selection affects all reported numbers.
  • LoFreq critical p-value threshold = 2^-200
    Taken from prior work [86], not fitted here.
assumptions (4)
  • domain assumption MPFR 256-bit results are exact for accuracy baselines.
    The paper treats 256-bit MPFR as ground truth. This is standard but unproven for the specific sums; error is not bounded.
  • domain assumption Relative error of the linear probability is the correct accuracy metric for the target applications.
    Assumed throughout Section IV and VI-D; not validated against downstream statistical decisions.
  • standard math The forward algorithm and PBD recurrences as implemented are the canonical forms.
    Standard definitions from literature cited in Section V-A.
  • domain assumption Posit arithmetic with MArTo is representative of posit hardware cost.
    The hardware results depend on MArTo implementation quality; authors note it is an HLS research prototype.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Design and accuracy trade-offs in Computational Statistics." pith.science (2026). https://pith.science/paper/LBFX44AG

@misc{pith2026250910934,
  author       = {Pith},
  title        = {Pith review of: Design and accuracy trade-offs in Computational Statistics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LBFX44AG}},
  note         = {Machine review of arXiv:2509.10934}
}
abstract

Statistical computations are becoming increasingly important. These computations often need to be performed in log-space because probabilities become extremely small due to repeated multiplications. While using logarithms effectively prevents numerical underflow, this paper shows that its cost is high in performance, resource utilization, and, notably, numerical accuracy. This paper then argues that using posit, a recently proposed floating-point format, is a better strategy for statistical computations operating on extremely small numbers because of its unique encoding mechanism. To that end, this paper performs a comprehensive analysis comparing posit, binary64, and logarithm representations, examining both individual arithmetic operations, statistical bioinformatics applications, and their accelerators. FPGA implementation results highlight that posit-based accelerators can achieve up to two orders of magnitude higher accuracy, up to 60\% lower resource utilization, and up to $1.3\times$ speedup, compared to log-space accelerators. Such improvement translates to $2\times$ performance per unit resource on the FPGA.

Figures

Figures reproduced from arXiv: 2509.10934 by the authors.

Figure 1
Figure 1. Base-2 exponent value of alpha over iterations. Section VI highlights the benefits of using posit. Section VII discusses related works and Section VIII concludes the paper. II. Problem A. Motivation Statistical computations are increasingly common in mod￾ern applications, where fundamental operations are the multi￾plication and addition of probabilities. In these applications, a key challenge is that numbers tend to… view at source ↗
Figure 2
Figure 2. IEEE Floating-Point and Posit Number Formats. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Individual operation accuracy on numbers of different magnitudes. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Processing Element (PE) in forward algorithm units. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Execution timeline of the accelerators. Besides, the posit-based accelerators consume less than half of the resources used by their logarithm-based counterparts. Such a resource-saving enables building more hardware units on the FPGA, leading to even greater speedup. M…
Figure 8
Figure 8. Figure 8: Performance per Resource Unit. As shown in [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]
Figure 9
Figure 9. Figure 9: This is why posit(64, 9) is absent in the two leftmost ranges of the figure, and binary64 is not shown at all. For most p-values, posit(64, 12) and posit(64, 9) have higher accuracy than using logarithms. In [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]
Figure 9
Figure 9. Figure 9: Accuracy of final p-values of different magnitudes. [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: Overall accuracy of final likelihoods in VICAR. [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]
Figure 11
Figure 11. Figure 11: Overall accuracy of final p-values in LoFreq. [PITH_FULL_IMAGE:figures/full_fig_p010_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

89 extracted references · 3 canonical work pages

  1. [1]

    [Online]

    The gnu mpfr library. [Online]. Available: https://www.mpfr.org/

  2. [2]

    Lofreq poisson binomial distribution implementation

    “Lofreq poisson binomial distribution implementation. ” [Online]. Avail- able: https://github.com/CSB5/lofreq/blob/master/src/lofreq/snpcaller.c# L821

  3. [3]

    [Online]

    Softmax classifier. [Online]. Available: https://cs231n.github.io/linear- classify/#softmax-classifier

  4. [4]

    Stan poisson binomial distribution implementation

    “Stan poisson binomial distribution implementation. ” [Online]. Available: https://github.com/stan-dev/math/blob/develop/stan/math/ prim/fun/poisson_binomial_log_probs.hpp

  5. [5]

    Ieee standard for floating-point arithmetic,

    “Ieee standard for floating-point arithmetic, ”IEEE Std 754-2019 (Revision of IEEE 754-2008), pp. 1–84, 2019

    2019

  6. [6]

    Low-precision logarithmic number systems: beyond base-2,

    S. A. Alam, J. Garland, and D. Gregg, “Low-precision logarithmic number systems: beyond base-2, ”ACM Transactions on Architecture and Code Optimization (TACO), vol. 18, no. 4, pp. 1–25, 2021

    2021

  7. [7]

    Sneakysnake: a fast and accurate universal genome pre-alignment filter for cpus, gpus and fpgas,

    M. Alser, T. Shahroodi, J. Gómez-Luna, C. Alkan, and O. Mutlu, “Sneakysnake: a fast and accurate universal genome pre-alignment filter for cpus, gpus and fpgas, ”Bioinformatics, vol. 36, no. 22-23, pp. 5282–5290, 2020

    2020

  8. [8]

    Acmc 2: Accelerating markov chain monte carlo algorithms for probabilistic models,

    S. S. Banerjee, Z. T. Kalbarczyk, and R. K. Iyer, “Acmc 2: Accelerating markov chain monte carlo algorithms for probabilistic models, ” in Proceedings of the Twenty-Fourth International Conference on Architectural Support for Programming Languages and Operating Systems, 2019, pp. 515–528

    2019

Show all 89 references
  1. [9]

    Efficient ancestry and mutation simulation with msprime 1.0,

    F. Baumdicker, G. Bisschop, D. Goldstein, G. Gower, A. P. Ragsdale, G. Tsambos, S. Zhu, B. Eldon, E. C. Ellerman, J. G. Galloway, A. L. Gladstein, G. Gorjanc, B. Guo, B. Jeffery, W. W. Kretzschumar, K. Lohse, M. Matschiner, D. Nelson, N. S. Pope, C. D. Quinto-Cortés, M. F. Rod...

    2021

  2. [10]

    Accurately computing the log-sum-exp and softmax functions,

    P. Blanchard, D. J. Higham, and N. J. Higham, “Accurately computing the log-sum-exp and softmax functions, ”IMA Journal of Numerical Analysis, vol. 41, no. 4, pp. 2311–2330, 08 2020. [Online]. Available: https://doi.org/10.1093/imanum/draa038

    2020 doi

  3. [11]

    Variational inference: A review for statisticians,

    D. M. Blei, A. Kucukelbir, and J. D. McAuliffe, “Variational inference: A review for statisticians, ”Journal of the American statistical Association, vol. 112, no. 518, pp. 859–877, 2017

    2017

  4. [12]

    Stan: A probabilistic programming language,

    B. Carpenter, A. Gelman, M. D. Hoffman, D. Lee, B. Goodrich, M. Betancourt, M. A. Brubaker, J. Guo, P. Li, and A. Riddell, “Stan: A probabilistic programming language, ”Journal of statistical software, vol. 76, 2017

    2017

  5. [13]

    Coopmc: Algorithm-architecture co-optimization for markov chain monte carlo accelerators,

    Y. Chai, G. G. Ko, W.-T. Mark Ting, L. Bailey, D. Brooks, and G.-Y. Wei, “Coopmc: Algorithm-architecture co-optimization for markov chain monte carlo accelerators, ” in2022 IEEE International Symposium on High-Performance Computer Architecture (HPCA), 2022, pp. 38–52

    2022

  6. [14]

    Parameterized posit arithmetic hardware generator,

    R. Chaurasiya, J. Gustafson, R. Shrestha, J. Neudorfer, S. Nambiar, K. Niyogi, F. Merchant, and R. Leupers, “Parameterized posit arithmetic hardware generator, ” in2018 IEEE 36th International Conference on Computer Design (ICCD). IEEE, 2018, pp. 334–341

    2018

  7. [15]

    Posit npb: Assessing the precision improvement in hpc scientific applications,

    S. W. Chien, I. B. Peng, and S. Markidis, “Posit npb: Assessing the precision improvement in hpc scientific applications, ” inInternational Conference on Parallel Processing and Applied Mathematics. Springer, 2019, pp. 301–310

    2019

  8. [16]

    Debugging and detecting numerical errors in computation with posits,

    S. Chowdhary, J. P. Lim, and S. Nagarakatte, “Debugging and detecting numerical errors in computation with posits, ” inProceedings of the 41st ACM SIGPLAN Conference on Programming Language Design and Implementation, ser. PLDI 2020. New York, NY, USA: Association for Computing...

    2020

  9. [17]

    Posits: the good, the bad and the ugly,

    F. De Dinechin, L. Forget, J.-M. Muller, and Y. Uguen, “Posits: the good, the bad and the ugly, ” inProceedings of the Conference for Next Generation Arithmetic 2019, 2019, pp. 1–10

    2019

  10. [18]

    Statistical significance: p value, 0.05 threshold, and applications to radiomics—reasons for a conservative approach,

    G. Di Leo and F. Sardanelli, “Statistical significance: p value, 0.05 threshold, and applications to radiomics—reasons for a conservative approach, ”European radiology experimental, vol. 4, no. 1, pp. 1–8, 2020

    2020

  11. [19]

    Durbin, S

    R. Durbin, S. R. Eddy, A. Krogh, and G. Mitchison,Biological sequence analysis: probabilistic models of proteins and nucleic acids. Cambridge university press, 1998

    1998

  12. [20]

    Hidden markov models,

    S. R. Eddy, “Hidden markov models, ”Current opinion in structural biology, vol. 6, no. 3, pp. 361–365, 1996

    1996

  13. [21]

    Accelerated profile hmm searches,

    ——, “Accelerated profile hmm searches, ”PLoS computational biology, vol. 7, no. 10, p. e1002195, 2011

    2011

  14. [22]

    High- performance fpga-based accelerator for bayesian neural networks,

    H. Fan, M. Ferianc, M. Rodrigues, H. Zhou, X. Niu, and W. Luk, “High- performance fpga-based accelerator for bayesian neural networks, ” in 2021 58th ACM/IEEE Design Automation Conference (DAC). IEEE, 2021, pp. 1063–1068

    2021

  15. [23]

    Closed-form expression for the poisson- binomial probability density function,

    M. Fernandez and S. Williams, “Closed-form expression for the poisson- binomial probability density function, ”IEEE Transactions on Aerospace and Electronic Systems, vol. 46, no. 2, pp. 803–817, 2010

    2010

  16. [24]

    Aphmm: Accelerating profile hidden markov models for fast and energy-efficient genome analysis,

    C. Firtina, K. Pillai, G. S. Kalsi, B. Suresh, D. Senol Cali, J. S. Kim, T. Shahroodi, M. B. Cavlak, J. Lindegger, M. Alser, J. Gómez Luna, S. Subramoney, and O. Mutlu, “Aphmm: Accelerating profile hidden markov models for fast and energy-efficient genome analysis, ”arXiv prep...

    2022 arXiv

  17. [25]

    Comparing posit and ieee-754 hardware cost,

    L. Forget, Y. Uguen, and F. de Dinechin, “Comparing posit and ieee-754 hardware cost, ” 2021

    2021

  18. [26]

    Mpfr: A multiple-precision binary floating-point library with correct rounding,

    L. Fousse, G. Hanrot, V. Lefèvre, P. Pélissier, and P. Zimmermann, “Mpfr: A multiple-precision binary floating-point library with correct rounding, ”ACM Trans. Math. Softw., vol. 33, no. 2, p. 13–es, jun 2007. [Online]. Available: https://doi.org/10.1145/1236463.1236468

    2007

  19. [27]

    Gendp: A framework of dynamic programming acceleration for genome sequencing analysis,

    Y. Gu, A. Subramaniyan, T. Dunn, A. Khadem, K.-Y. Chen, S. Paul, M. Vasimuddin, S. Misra, D. Blaauw, S. Narayanasamy, and R. Das, “Gendp: A framework of dynamic programming acceleration for genome sequencing analysis, ” inProceedings of the 50th Annual International Symposium ...

    2023

  20. [28]

    Gundersen

    G. Gundersen. (2020) The log-sum-exp trick. [Online]. Available: https://gregorygundersen.com/blog/2020/02/09/log-sum-exp/

    2020

  21. [29]

    Hardware acceleration of long read pairwise overlapping in genome sequencing: A race between fpga and gpu,

    L. Guo, J. Lau, Z. Ruan, P. Wei, and J. Cong, “Hardware acceleration of long read pairwise overlapping in genome sequencing: A race between fpga and gpu, ” in2019 IEEE 27th Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM). IEEE, 2019, pp. 127–135

    2019

  22. [30]

    Beating floating point at its own game: Posit arithmetic,

    J. L. Gustafson and I. T. Yonemoto, “Beating floating point at its own game: Posit arithmetic, ”Supercomputing frontiers and innovations, vol. 4, no. 2, pp. 71–86, 2017

    2017

  23. [31]

    Bridging the gap between llms and lns with dynamic data format and architecture codesign,

    P. Haghi, C. Wu, Z. Azad, Y. Li, A. Gui, Y. Hao, A. Li, and T. T. Geng, “Bridging the gap between llms and lns with dynamic data format and architecture codesign, ” in2024 57th IEEE/ACM International Symposium on Microarchitecture (MICRO). IEEE, 2024, pp. 1617–1631

    2024

  24. [32]

    On computing the distribution function for the poisson binomial distribution,

    Y. Hong, “On computing the distribution function for the poisson binomial distribution, ”Computational Statistics & Data Analysis, vol. 59, pp. 41–51, 2013

    2013

  25. [33]

    Hardware acceleration of the pair-hmm algorithm for dna variant calling,

    S. Huang, G. J. Manikandan, A. Ramachandran, K. Rupnow, W.- m. W. Hwu, and D. Chen, “Hardware acceleration of the pair-hmm algorithm for dna variant calling, ” inProceedings of the 2017 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays, 2017, pp. 275–284

    2017

  26. [34]

    Universal number posit arithmetic generator on fpga,

    M. K. Jaiswal and H. K.-H. So, “Universal number posit arithmetic generator on fpga, ” in2018 Design, Automation Test in Europe Conference Exhibition (DATE), 2018, pp. 1159–1162

    2018

  27. [35]

    Pacogen: A hardware posit arithmetic core generator,

    ——, “Pacogen: A hardware posit arithmetic core generator, ”IEEE access, vol. 7, pp. 74 586–74 601, 2019

    2019

  28. [36]

    Exma: A genomics accelerator for exact- matching,

    L. Jiang and F. Zokaee, “Exma: A genomics accelerator for exact- matching, ” in2021 IEEE International Symposium on High-Performance Computer Architecture (HPCA). IEEE, 2021, pp. 399–411

    2021

  29. [37]

    Rethinking floating point for deep learning,

    J. Johnson, “Rethinking floating point for deep learning, ”arXiv preprint arXiv:1811.01721, 2018

    2018 arXiv

  30. [38]

    Jurafsky and J

    D. Jurafsky and J. H. Martin. (2023) Speech and language processing. chapter 3: N-gram language models. [Online]. Available: https://web.stanford.edu/~jurafsky/slp3/3.pdf

    2023

  31. [39]

    Towards weather and climate models in 16-bit arithmetic

    M. Klöwer, “Towards weather and climate models in 16-bit arithmetic. ”

  32. [40]

    A scalable bayesian inference accelerator for unsupervised learning,

    G. Ko, Y. Chai, M. Donato, P. N. Whatmough, T. Tambe, R. A. Rutenbar, G.-Y. Wei, and D. Brooks, “A scalable bayesian inference accelerator for unsupervised learning, ” in2020 IEEE Hot Chips 32 Symposium (HCS). IEEE Computer Society, 2020, pp. 1–27

    2020

  33. [41]

    A 3mm 2 programmable bayesian inference accelerator for unsupervised machine perception using parallel gibbs sampling in 16nm,

    G. G. Ko, Y. Chai, M. Donato, P. N. Whatmough, T. Tambe, R. A. Rutenbar, D. Brooks, and G.-Y. Wei, “A 3mm 2 programmable bayesian inference accelerator for unsupervised machine perception using parallel gibbs sampling in 16nm, ” in2020 IEEE Symposium on VLSI Circuits. IEEE, 20...

    2020

  34. [42]

    Flexgibbs: Reconfigurable parallel gibbs sampling accelerator for structured graphs,

    G. G. Ko, Y. Chai, R. A. Rutenbar, D. Brooks, and G.-Y. Wei, “Flexgibbs: Reconfigurable parallel gibbs sampling accelerator for structured graphs, ” in2019 IEEE 27th Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM). IEEE, 2019, pp. 334–334. 11

    2019

  35. [43]

    High-level. net software implementations of unum type i and posit with simultaneous fpga implementation using hastlayer,

    Z. Lehóczky, A. Retzler, R. Tóth, Á. Szabó, B. Farkas, and K. Somogyi, “High-level. net software implementations of unum type i and posit with simultaneous fpga implementation using hastlayer, ” inProceedings of the Conference for Next Generation Arithmetic, 2018, pp. 1–7

    2018

  36. [44]

    A robust scaling approach for implementation of hsmms,

    B.-C. Li and S.-Z. Yu, “A robust scaling approach for implementation of hsmms, ”IEEE Signal Processing Letters, vol. 22, no. 9, pp. 1264–1268, 2015

    2015

  37. [45]

    An approach to generate correctly rounded math libraries for new floating point variants,

    J. P. Lim, M. Aanjaneya, J. Gustafson, and S. Nagarakatte, “An approach to generate correctly rounded math libraries for new floating point variants, ”Proceedings of the ACM on Programming Languages, vol. 5, no. POPL, pp. 1–30, 2021

    2021

  38. [46]

    High performance correctly rounded math libraries for 32-bit floating point representations,

    J. P. Lim and S. Nagarakatte, “High performance correctly rounded math libraries for 32-bit floating point representations, ” inProceedings of the 42nd ACM SIGPLAN International Conference on Programming Language Design and Implementation, 2021, pp. 359–374

    2021

  39. [47]

    Variational inference using approximate likelihood under the coalescent with recombination,

    X. Liu, H. A. Ogilvie, and L. Nakhleh, “Variational inference using approximate likelihood under the coalescent with recombination, ” Genome Research, vol. 31, no. 11, pp. 2107–2119, 2021

    2021

  40. [48]

    Algorithm-hardware co-design for bqsr acceleration in genome analysis toolkit,

    M. Lo, Z. Fang, J. Wang, P. Zhou, M.-C. F. Chang, and J. Cong, “Algorithm-hardware co-design for bqsr acceleration in genome analysis toolkit, ” in2020 IEEE 28th Annual International Symposium on Field- Programmable Custom Computing Machines (FCCM). IEEE, 2020, pp. 157–166

    2020

  41. [49]

    Big-percival: Exploring the native use of 64-bit posit arithmetic in scientific computing,

    D. Mallasén, A. A. Del Barrio, and M. Prieto-Matias, “Big-percival: Exploring the native use of 64-bit posit arithmetic in scientific computing, ”arXiv preprint arXiv:2305.06946, 2023

    2023 arXiv

  42. [50]

    Customizing the cva6 risc-v core to integrate posit and quire instructions,

    D. Mallasén, R. Murillo, A. A. Del Barrio, G. Botella, L. Piñuel, and M. Prieto-Matias, “Customizing the cva6 risc-v core to integrate posit and quire instructions, ” in2022 37th Conference on Design of Circuits and Integrated Circuits (DCIS). IEEE, 2022, pp. 01–06

    2022

  43. [51]

    Percival: open-source posit risc-v core with quire capability,

    ——, “Percival: open-source posit risc-v core with quire capability, ”IEEE Transactions on Emerging Topics in Computing, vol. 10, no. 3, pp. 1241– 1252, 2022

    2022

  44. [52]

    Numerically stable hidden markov model implementation,

    T. P. Mann, “Numerically stable hidden markov model implementation, ” An HMM scaling tutorial, pp. 1–8, 2006

    2006

  45. [53]

    A systematic review of hidden markov models and their applications,

    B. Mor, S. Garhwal, and A. Kumar, “A systematic review of hidden markov models and their applications, ”Archives of computational methods in engineering, vol. 28, pp. 1429–1448, 2021

    2021

  46. [54]

    Customized posit adders and multipliers using the flopoco core generator,

    R. Murillo, A. A. Del Barrio, and G. Botella, “Customized posit adders and multipliers using the flopoco core generator, ” in2020 IEEE International Symposium on Circuits and Systems (ISCAS). IEEE, 2020, pp. 1–5

    2020

  47. [55]

    Generating posit-based accelerators with high-level synthesis,

    R. Murillo, A. A. Del Barrio, G. Botella, and C. Pilato, “Generating posit-based accelerators with high-level synthesis, ”IEEE Transactions on Circuits and Systems I: Regular Papers, 2023

    2023

  48. [56]

    Compar- ing different decodings for posit arithmetic,

    R. Murillo, D. Mallasén, A. A. Del Barrio, and G. Botella, “Compar- ing different decodings for posit arithmetic, ” inConference on Next Generation Arithmetic. Springer, 2022, pp. 84–99

    2022

  49. [57]

    Plaus: Posit logarithmic approximate units to implement low- cost operations with real numbers,

    ——, “Plaus: Posit logarithmic approximate units to implement low- cost operations with real numbers, ” inConference on Next Generation Arithmetic. Springer, 2023, pp. 171–188

    2023

  50. [58]

    Hidden semi-markov models (hsmms),

    K. P. Murphy, “Hidden semi-markov models (hsmms), ”unpublished notes, vol. 2, 2002

    2002

  51. [59]

    AWTY (are we there yet?): a system for graphical exploration of MCMC convergence in Bayesian phylogenetics,

    J. A. Nylander, J. C. Wilgenbusch, D. L. Warren, and D. L. Swofford, “AWTY (are we there yet?): a system for graphical exploration of MCMC convergence in Bayesian phylogenetics, ” Bioinformatics, vol. 24, no. 4, pp. 581–583, 08 2007. [Online]. Available: https://doi.org/10.109...

    2007 doi

  52. [60]

    Universal: reliable, reproducible, and energy-efficient numerics,

    E. T. L. Omtzigt and J. Quinlan, “Universal: reliable, reproducible, and energy-efficient numerics, ” inConference on Next Generation Arithmetic. Springer, 2022, pp. 100–116

    2022

  53. [61]

    Computing with logarithmic number system arithmetic: Implementation methods and performance benefits,

    B. Parhami, “Computing with logarithmic number system arithmetic: Implementation methods and performance benefits, ”Computers & Electrical Engineering, vol. 87, p. 106800, 2020

    2020

  54. [62]

    C. Piech. (2023) Stanford’s cs109 course, probability for computer scientists. [Online]. Available: https://chrispiech.github. io/probabilityForComputerScientists/en/part1/log_probabilities/

    2023

  55. [63]

    Hardware implementation of posits and their application in fpgas,

    A. Podobas and S. Matsuoka, “Hardware implementation of posits and their application in fpgas, ” in2018 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW). IEEE, 2018, pp. 138–145

    2018

  56. [64]

    A tutorial on hidden markov models and selected applica- tions in speech recognition,

    L. Rabiner, “A tutorial on hidden markov models and selected applica- tions in speech recognition, ”Proceedings of the IEEE, vol. 77, no. 2, pp. 257–286, 1989

    1989

  57. [65]

    Some new results for poisson binomial models,

    E. Rosenman, “Some new results for poisson binomial models, ”arXiv preprint arXiv:1907.09053, 2019

    1907 arXiv

  58. [66]

    Fpga acceleration of sequence alignment: A survey,

    S. Salamat and T. Simunic, “Fpga acceleration of sequence alignment: A survey, ”ArXiv, vol. abs/2002.02394, 2020

    2002 arXiv

  59. [67]

    Techniques for assessing water infrastructure for nonstationary extreme events: a review,

    J. Salas, J. Obeysekera, and R. Vogel, “Techniques for assessing water infrastructure for nonstationary extreme events: a review, ”Hydrological Sciences Journal, vol. 63, no. 3, pp. 325–352, 2018

    2018

  60. [68]

    Binomial option pricing with nonidentically distributed returns and its implications,

    N. Schumacher, “Binomial option pricing with nonidentically distributed returns and its implications, ”Mathematical and computer modelling, vol. 29, no. 10-12, pp. 121–143, 1999

    1999

  61. [69]

    Clarinet: A quire-enabled risc-v- based framework for posit arithmetic empiricism,

    N. N. Sharma, R. Jain, M. M. Pokkuluri, S. B. Patkar, R. Leupers, R. S. Nikhil, and F. Merchant, “Clarinet: A quire-enabled risc-v- based framework for posit arithmetic empiricism, ”Journal of Systems Architecture, vol. 135, p. 102801, 2023

    2023

  62. [70]

    Phac: Posit hardware accelerator for efficient arithmetic logic operations,

    D. Shekhawat, J. Gandhi, M. Santosh, and J. G. Pandey, “Phac: Posit hardware accelerator for efficient arithmetic logic operations, ” in Conference on Next Generation Arithmetic. Springer, 2023, pp. 88– 100

    2023

  63. [71]

    Comparison of arithmetic number formats for inference in sum-product networks on fpgas,

    L. Sommer, L. Weber, M. Kumm, and A. Koch, “Comparison of arithmetic number formats for inference in sum-product networks on fpgas, ” in 2020 IEEE 28th Annual international symposium on field-programmable custom computing machines (FCCM). IEEE, 2020, pp. 75–83

    2020

  64. [72]

    Accelerated seeding for genome sequence alignment with enumerated radix trees,

    A. Subramaniyan, J. Wadden, K. Goliya, N. Ozog, X. Wu, S. Narayanasamy, D. Blaauw, and R. Das, “Accelerated seeding for genome sequence alignment with enumerated radix trees, ” in2021 ACM/IEEE 48th Annual International Symposium on Computer Architec- ture (ISCA). IEEE, 2021, p...

    2021

  65. [73]

    The poisson binomial distribution—old & new,

    W. Tang and F. Tang, “The poisson binomial distribution—old & new, ” Statistical Science, vol. 38, no. 1, pp. 108–119, 2023

    2023

  66. [74]

    Stan user’s guide, version 2.35,

    S. D. Team, “Stan user’s guide, version 2.35, ” 2024. [Online]. Available: https://mc-stan.org/docs/stan-users-guide/floating- point.html#arithmetic-precision

    2024

  67. [75]

    The role of poisson’s binomial distribution in the analysis of tem images,

    A. Tejada and A. J. den Dekker, “The role of poisson’s binomial distribution in the analysis of tem images, ”Ultramicroscopy, vol. 111, no. 11, pp. 1553–1556, 2011

    2011

  68. [76]

    Peri: A configurable posit enabled risc-v core,

    S. Tiwari, N. Gala, C. Rebeiro, and V. Kamakoti, “Peri: A configurable posit enabled risc-v core, ”ACM Transactions on Architecture and Code Optimization (TACO), vol. 18, no. 3, pp. 1–26, 2021

    2021

  69. [77]

    Darwin: A genomics co- processor provides up to 15,000 x acceleration on long read assembly,

    Y. Turakhia, G. Bejerano, and W. J. Dally, “Darwin: A genomics co- processor provides up to 15,000 x acceleration on long read assembly, ” ACM SIGPLAN Notices, vol. 53, no. 2, pp. 199–213, 2018

    2018

  70. [78]

    Evaluating the hardware cost of the posit number system,

    Y. Uguen, L. Forget, and F. de Dinechin, “Evaluating the hardware cost of the posit number system, ” in2019 29th International Conference on Field Programmable Logic and Applications (FPL). IEEE, 2019, pp. 106–113

    2019

  71. [79]

    Enabling high performance posit arithmetic applications using hardware acceleration,

    L. van Dam, “Enabling high performance posit arithmetic applications using hardware acceleration, ” 2018

    2018

  72. [80]

    An accelerator for posit arithmetic targeting posit level 1 blas routines and pair-hmm,

    L. Van Dam, J. Peltenburg, Z. Al-Ars, and H. P. Hofstee, “An accelerator for posit arithmetic targeting posit level 1 blas routines and pair-hmm, ” inProceedings of the Conference for Next Generation Arithmetic 2019, 2019, pp. 1–10

    2019

  73. [81]

    Coestimating Reticulate Phylogenies and Gene Trees from Multilocus Sequence Data,

    D. Wen and L. Nakhleh, “Coestimating Reticulate Phylogenies and Gene Trees from Multilocus Sequence Data, ”Systematic Biology, vol. 67, no. 3, pp. 439–457, 10 2017. [Online]. Available: https://doi.org/10.1093/sysbio/syx085

    2017 doi

  74. [82]

    Lofreq: a sequence-quality aware, ultra-sensitive variant caller for uncovering cell- population heterogeneity from high-throughput sequencing datasets,

    A. Wilm, P. P. K. Aw, D. Bertrand, G. H. T. Yeo, S. H. Ong, C. H. Wong, C. C. Khor, R. Petric, M. L. Hibberd, and N. Nagarajan, “Lofreq: a sequence-quality aware, ultra-sensitive variant caller for uncovering cell- population heterogeneity from high-throughput sequencing datas...

    2012

  75. [83]

    Posit arithmetic hardware implementations with the minimum cost divider and squareroot,

    F. Xiao, F. Liang, B. Wu, J. Liang, S. Cheng, and G. Zhang, “Posit arithmetic hardware implementations with the minimum cost divider and squareroot, ”Electronics, vol. 9, no. 10, 2020. [Online]. Available: https://www.mdpi.com/2079-9292/9/10/1622

    2020

  76. [84]

    (2017) Ultrascale architecture configurable logic block user guide

    Xilinx. (2017) Ultrascale architecture configurable logic block user guide. [Online]. Available: https://docs.xilinx.com/v/u/en-US/ug574- ultrascale-clb

    2017

  77. [85]

    (2020) Floating-point operator v7.1 - logicore ip product guide

    ——. (2020) Floating-point operator v7.1 - logicore ip product guide. [Online]. Available: https://docs.xilinx.com/v/u/en-US/pg060-floating- point

    2020

  78. [86]

    An fpga accelerator for genome variant calling,

    T. Xu, S. Rixner, and A. L. Cox, “An fpga accelerator for genome variant calling, ”ACM Trans. Reconfigurable Technol. Syst., vol. 16, no. 4, sep 2023. [Online]. Available: https://doi.org/10.1145/3595297 12

    2023 doi

  79. [87]

    Accelerating probabilistic computing with a stochastic processing unit,

    X. Zhang, “Accelerating probabilistic computing with a stochastic processing unit, ” Ph.D. dissertation, Duke University, 2020

    2020

  80. [88]

    Statistical robustness of markov chain monte carlo accelerators,

    X. Zhang, R. Bashizade, Y. Wang, S. Mukherjee, and A. R. Lebeck, “Statistical robustness of markov chain monte carlo accelerators, ” in Proceedings of the 26th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, 2021, pp. 959–974

    2021

  81. [89]

    Lns-madam: Low-precision training in logarithmic number system using multiplicative weight update,

    J. Zhao, S. Dai, R. Venkatesan, B. Zimmer, M. Ali, M.-Y. Liu, B. Khailany, W. J. Dally, and A. Anandkumar, “Lns-madam: Low-precision training in logarithmic number system using multiplicative weight update, ”IEEE Transactions on Computers, vol. 71, no. 12, pp. 3179–3190, 2022. 13

    2022

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.