Pith. sign in

REVIEW 3 major objections 5 minor 43 references

SODECL: An Open Source Library for Calculating Multiple Orbits of a System of Stochastic Differential Equations in Parallel

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read SODECL computes many SDE orbits in parallel on CPUs and GPUs, reaching up to ~6.7x speedup on 32 cores and ~4.5x on a GPU.

desk verdict SODECL is a real, openly licensed library with honest modest speedups, but the headline speedup ratios rest on an under-specified single-core baseline that needs clarification before they are quoted. read the letter →

arxiv 1908.03869 v2 pith:AXIEHAKP submitted 2019-08-11 cs.MS

Eleftherios Avramidis , Marta Lalik , Ozgur E. Akman This is my paper · ORCID
classification cs.MS MSC 65C3065Y05
keywords stochasticdifferentialequationsEuler-MaruyamamethodparallelcomputingOpenCLgraphicsprocessingunitsmulti-coreCPUsKuramotomodel
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

The paper introduces SODECL, an open-source C++ library that computes many independent trajectories—'orbits'—of a system of stochastic differential equations at once by running the Euler-Maruyama method inside an OpenCL kernel that can execute on a multi-core CPU or a GPU. The claim is that this embarrassingly parallel workload, one of the main computational bottlenecks in fitting noisy dynamical models to data, can be moved to a single workstation or HPC node and accelerated substantially. On a stochastic Kuramoto model, the library achieved up to about 6.7-fold speedup using all 32 cores of a high-end HPC node and about 4.5-fold on a high-end GPU, relative to a single core, and it was faster than a parallel MATLAB implementation in almost all tested configurations. The paper also shows that the solver reproduces the known Kuramoto synchronization transition, supporting the numerical reliability of the parallel implementation. If the speedups generalize to other SDE systems, parameter optimisation and uncertainty analysis for stochastic biological and physical models become practical on ordinary parallel hardware.

What carries the argument

The central mechanism is a single OpenCL kernel, named integrator_caller, generated at runtime from the model definition. Each kernel call evolves every orbit for a fixed number of Euler-Maruyama steps, with a counter-based random-number generator supplying independent normal deviates through the Box-Muller transform. The user supplies two functions—the drift and the diffusion of the SDE—and the library builds the kernel source string, so no recompilation is needed when the model changes.

What would settle it

Run the same orbit-batch benchmark on a second SDE model—for example a stochastic gene-regulatory or neuron model—on the same 32-core node and P100 GPU using the paper's hand-tuned local group sizes; if speedups at comparable orbit counts fall far below the reported 6.7x and 4.5x, the acceleration is a feature of the benchmark model rather than the library's parallel design.

Watch

Extended reading notes

Core claim

The central claim is that replacing the single-threaded, orbit-by-orbit loop with an OpenCL kernel that assigns one work item (or work group) to each orbit gives substantial, repeatable speedups on both multi-core CPUs and GPUs, without sacrificing numerical reliability. On the stochastic Kuramoto model, speedups relative to one logical core of a 32-core Intel Xeon Gold 6142 node reached about 6.7 with all 32 CPU cores (ranging from 1.12 to 6.65 depending on oscillator count and orbit number) and about 4.5 on an NVIDIA Tesla P100 GPU (ranging from 1.57 to 4.50). The GPUs were generally slower than all 32 CPU cores but noticeably faster than a high-end desktop CPU, and SODECL was faster than a parallel MATLAB implementation in almost every configuration, up to 5.8-fold. The numerical tests show that the solver reproduces the Kuramoto transition: phase coherence stays near zero below the critical coupling strength and saturates near one above it, with weak dependence on the integration time step.

Load-bearing premise

The load-bearing premise is that speedups measured on one stochastic Kuramoto model, with local group sizes hand-tuned for each device, represent what SODECL will deliver on other SDE systems and hardware; the paper itself lists benchmarks on other canonical models as a needed follow-up.

Editorial extensions

If this is right

  • At batch sizes of tens to hundreds of thousands of orbits, a single multi-core node or GPU can replace a cluster for the integration step of SDE parameter optimisation.
  • The speedup is configuration dependent: small orbit counts and small model sizes can run slower than one core (speedups below 1), so the benefit appears when the workload is large enough to amortise kernel and device overheads.
  • Because one OpenCL kernel targets CPUs and GPUs from multiple vendors, the same library code can be deployed on workstations, GPU nodes, and CPU-only HPC nodes without hardware-specific rewrites.
  • The accuracy tests tie the speed measurement to a physically meaningful outcome: the parallel Euler-Maruyama solver reproduces the expected Kuramoto coherence transition, with observables only weakly dependent on the integration time step.

Reading between the lines

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

  • If the speedups transfer to other models, an automatic tuning step for the OpenCL local group size would strengthen the library, since the reported numbers rely on hand-picked values per device and model size.
  • The fact that 32 CPU cores yield only about 6.7x rather than anything near 32x suggests memory bandwidth and kernel-launch overhead limit this workload, so splitting the integration across multiple kernels or reducing host-device transfers could yield further gains.
  • Because double precision was used throughout and GPUs typically accelerate single precision considerably more, the GPU speedups reported here may be a lower bound for applications that can tolerate lower numerical precision.
  • A natural benchmark extension would be a second stochastic model with heterogeneous equations and branchy terms, which the authors expect to vectorise less well and could expose whether the acceleration is model-specific.
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 manuscript presents SODECL, an open-source C++/OpenCL library for computing many independent orbits of an SDE/ODE system in parallel on CPUs and GPUs. The authors benchmark a stochastic Kuramoto model with N = 5, 10, and 15 oscillators and orbit counts from 512 to 163,840, measuring runtimes on a desktop CPU (i7-4790K), a two-socket Xeon Gold 6142 node, and two GPUs (FirePro W8100, Tesla P100). They report speedups up to about 6.7 for 32 CPU cores and about 4.5 for the P100 GPU relative to a single core of the Xeon node, compare against a MATLAB/OpenMP implementation, and verify numerical stability via phase-coherence statistics across timesteps for N = 100. The paper is primarily a software and benchmarking contribution.

Significance. The contribution is practical: a portable, open-source library for embarrassingly parallel SDE simulation, rather than a new numerical method. If the speedup measurements are robust, SODECL is a useful tool for parameter-sweep and population-based optimization workflows. Strengths of the paper include the public MIT-licensed code, multi-OS support, the detailed supplementary tables of raw runtimes, and the authors' explicit list of limitations in Section 3 (no initialization-time measurements, no single-precision tests, no branch-divergence study). The reported speedups are modest and plausible for the workload, and the numerical stability checks in Section 2.3 provide a reasonable sanity check. The main weaknesses are in the benchmark methodology, not in the library's design.

major comments (3)
  1. [Section 2.2.1, Section 2.2.2, Figure 7, Tables S1-S6] The speedup baseline is defined inconsistently: Section 2.2.1 says 'only one core' of the HPC CPU, while Section 2.2.2 and the Figure 7 caption state that speedups are relative to 'one logical processor' of the 2X6142. The Intel Xeon Gold 6142 has two hardware threads per physical core and the node has two sockets, so a 'logical processor' baseline can differ materially from a physical-core baseline, depending on whether the sibling hyperthread is idle or active. The paper also does not state whether the single-core baseline used the local group size of 32 from Table 4 or a group size of 1, nor whether hyperthreading was disabled. Since every headline speedup (e.g., 6.652 in Table S6 and 4.503 in Table S6) is the single-core 2X6142 runtime divided by the device runtime, this ambiguity directly affects the central quantitative claims. The accuracy tests in Section 2.3 cannot resolve this issue because they test numerical stability, not the timing baseline. Please specify the exact baseline configuration and report runtimes for a dedicated physical-core baseline as well.
  2. [Section 2.2.1, Tables S1-S6, Figure 7] All runtimes are reported as means of eight runs, but no standard deviation, min-max range, or per-run values are given. Without a measure of dispersion, the reader cannot assess whether differences across orbit counts and devices (for example, the 2X6142 speedups of 4.738 and 4.791 in Table S5 at 40960 and 81920 orbits) are within run-to-run noise, nor whether the reported maxima are stable. For an empirical benchmark paper whose central claim is a set of quantitative speedups, reporting variance is necessary. Please add error bars to Figure 7 and dispersion columns to the supplementary tables, and state whether the timed loop included OpenCL kernel compilation and device initialization.
  3. [Table 4, Section 2.2.2, Eq. (5), Section 3] The reported speedups were obtained with local group sizes manually tuned for each device and model size (Table 4), and with a single canonical SDE model, the stochastic Kuramoto system (Eq. 5). The authors acknowledge in Section 3 that further benchmarks on other models are needed, but the abstract and the beginning of Section 2.2.2 present the maximum speedups without this qualifier. Because the headline numbers are the result of per-device, per-model-size tuning, they should be presented as 'speedups for the tested configurations with tuned local group sizes,' or the paper should include a sensitivity analysis showing how the speedup varies with local group size (e.g., group sizes 1, 8, 32, 64, 128) for at least one device and model size.
minor comments (5)
  1. [Table S5] The column header 'I4I90K' is a typo for 'I4790K'.
  2. [Title] The title contains 'A N OPEN SOURCE LIBRARY'; this should read 'AN OPEN SOURCE LIBRARY'.
  3. [Throughout] The text renders 'MATLAB' with a space as 'MA TLAB' in several places (e.g., Section 2.2.1 and Figure 6); please fix the spacing.
  4. [Figure 8 and Tables S7-S9] Figure 8 says runtimes are on 'all logical processors of the I4790K CPU,' while Tables S7-S9 say 'all cores'; align these terms with the clarification requested for the speedup baseline.
  5. [Figure 6] The MATLAB code hard-codes a parfor pool of 8 workers; because the i7-4790K has 4 physical cores and 8 logical processors, the comparison should state whether MATLAB used 8 software threads on logical processors and whether SODECL used the same number of threads.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the paper is an empirical performance benchmark whose speedups are direct runtime measurements, not derived quantities or fitted predictions.

full rationale

SODECL reports measured runtimes and speedups for a stochastic Kuramoto model across CPUs and GPUs. The claimed results, such as a maximum speedup of ~6.7 on 32 CPU cores and ~4.5 on a GPU, are ratios of directly observed single-core and multi-core/GPU runtimes (Tables S1-S3, S4-S6), not quantities defined in terms of the conclusions they support. There is no fitted parameter later renamed as a prediction, no ansatz imported via citation, and no uniqueness theorem invoked to force the choice of method. The paper's self-citations, particularly the comparison with the earlier speedup of ~20 in [16], explain a discrepancy but do not constitute the evidence for the new measurements. Concerns about whether the single-core baseline was a logical processor or a physical core, or whether the manually tuned local group sizes bias the speedups, are legitimate empirical-benchmarking caveats and are even partially acknowledged by the authors, but they are not circularity: the speedups remain externally measurable observations rather than the outputs of a derivation that assumes them. The accuracy and stability checks are independent numerical validations of the solver, and they do not determine the timing results. The paper is therefore self-contained as an empirical benchmark, and the appropriate circularity score is 0.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The central claim is an empirical performance benchmark. It depends on one set of tuned free parameters, the local group sizes, and on domain assumptions about the representativeness of the test model and the portability of OpenCL. No new theoretical entities are introduced.

free parameters (1)
  • Local group sizes per device and model size = I4790K: 8, 8, 8; 2X6142: 32, 32, 32; W8100: 256, 16, 32; P100: 8, 8, 8 for N = 5, 10, 15
    Chosen by exploratory tests because no analytical formula is known. The reported speedups depend on these values.
assumptions (4)
  • standard math The Euler-Maruyama method converges to the Itô solution of the SDE as the timestep tends to zero.
    This is the numerical scheme in Eq. (2), and the accuracy tests in Section 2.3 assume it is a valid integrator.
  • domain assumption The stochastic Kuramoto model in Eq. (5) is representative of the SDE models SODECL is intended to support.
    All performance benchmarks use this single model, and the authors note that other canonical models should be tested.
  • domain assumption OpenCL gives portably good performance across the tested CPUs and GPUs.
    The library is built around OpenCL for hardware portability, and the benchmark assumes the results reflect this portability.
  • ad hoc to paper The manually tuned local group sizes are appropriate or near-optimal for the tested configurations.
    The values in Table 4 were found by exploratory tests, and the speedups are conditional on these choices.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SODECL: An Open Source Library for Calculating Multiple Orbits of a System of Stochastic Differential Equations in Parallel." pith.science (2026). https://pith.science/paper/AXIEHAKP

@misc{pith2026190803869,
  author       = {Pith},
  title        = {Pith review of: SODECL: An Open Source Library for Calculating Multiple Orbits of a System of Stochastic Differential Equations in Parallel},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AXIEHAKP}},
  note         = {Machine review of arXiv:1908.03869}
}
abstract

Stochastic differential equations (SDEs) are widely used to model systems affected by random processes. In general, the analysis of an SDE model requires numerical solutions to be generated many times over multiple parameter combinations. However, this process often requires considerable computational resources to be practicable. Due to the embarrassingly parallel nature of the task, devices such as multi-core processors and graphics processing units (GPUs) can be employed for acceleration. Here, we present {\bf SODECL} (\url{https://github.com/avramidis/sodecl}), a software library that utilises such devices to calculate multiple orbits of an SDE model. To evaluate the acceleration provided by SODECL, we compared the time required to calculate multiple orbits of an exemplar stochastic model when one CPU core is used, to the time required when using all CPU cores or a GPU. In addition, to assess scalability, we investigated how the model size affected execution time on different parallel compute devices. Our results show that when using all 32 CPU cores of a high-end high-performance computing node, the task is accelerated by a factor of up to $\simeq$6.7, compared to when using a single CPU core. Executing the task on a high-end GPU yielded accelerations of up to $\simeq$4.5, compared to a single CPU core.

Figures

Figures reproduced from arXiv: 1908.03869 by the authors.

Figure 1
Figure 1. Definition of the SODECL kernel function. The funct [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Definition of the SODECL system_integrator function. The function parameters are the current time value of the SDE system (t), the corresponding values of the dependent variables (y), the parameter values (p), the current noise values (noise) and the new values of the dependent variables calculated by the Euler-Maruyama method (yout). void s o d e _ s y st em ( double t , double y [ _numeq_ ] , double y o ut [ _nume… view at source ↗
Figure 3
Figure 3. Definition of the SODECL functions specifying the S [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Definition of the SODECL Python wrapper function. T [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: SODECL implementation of the stochastic Kuramoto [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: MATLAB function for calculating orbits of the stoc [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Runtimes in seconds (s) and speedups for the stocha [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]
Figure 8
Figure 8. Figure 8: Runtime as a function of the number of orbits for the [PITH_FULL_IMAGE:figures/full_fig_p011_8.png]
Figure 9
Figure 9. Figure 9: Temporal evolution of phase coherence r(t) in the stochastic Kuramoto model (5) when N = 100 for K = 0.02 (black lines) and K = 0.2 (red lines). hr(t)i and σ (r(t)) denote the mean and standard deviation of r(t), respectively, calculated from 64 independent realisation…
Figure 10
Figure 10. Figure 10: Temporal evolution of oscillator phases θi(t) in the stochastic Kuramoto model. Each kymograph corre￾sponds to one of the realisations of (6) used to compute the phase coherence statistics shown in [PITH_FULL_IMAGE:figures/full_fig_p013_10.png]
Figure 11
Figure 11. Figure 11: Dependence on the integration timestep ∆t of the phase coherence statistics hr(t)i and σ (r(t)) at the end of the integration interval tMAX = 400. Black lines and axes show results for K = 0.02; red lines and axes show results for K = 0.2. In each case, hr(tMAX )i and…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 43 canonical work pages

  1. [1]

    Elston, William J

    Mads Kaern, Timothy C. Elston, William J. Blake, and Jame s J. Collins. Stochasticity in gene expression: from theories to phenotypes. Nat. Rev. Genet., 6(6):451–464, 2005

    work page 2005

  2. [2]

    Noise in gene expression is coupled to growth rate

    Leeat Keren, David V an Dijk, Shira Weingarten-Gabbay, D an Davidi, Ghil Jona, Adina Weinberger, Ron Milo, and Eran Segal. Noise in gene expression is coupled to growth rate. Genome Res., 25(12):1893–1902, 2015

    work page 1902

  3. [3]

    Aldo Faisal, Luc P

    A. Aldo Faisal, Luc P . J. Selen, and Daniel M. Wolpert. Noi se in the nervous system. Nat. Rev. Neurosci., 9(4):292–303, 2008

    work page 2008

  4. [4]

    McIntosh, Olaf Sp orns, and Rolf Kötter

    Gustavo Deco, Viktor Jirsa, Anthony R. McIntosh, Olaf Sp orns, and Rolf Kötter. Key role of coupling, delay, and noise in resting brain fluctuations. Proc. Natl. Acad. Sci. U. S. A., 106(25):10302–10307, 2009

    work page 2009

  5. [5]

    Hooman Darabi and Asad A. Abidi. Noise in rf-cmos mixers: A simple physical model. IEEE J. Solid-State Circuits, 35(1):15–25, 2000

    work page 2000

  6. [6]

    Scholten, Luuk F

    Andries J. Scholten, Luuk F. Tiemeijer, Ronald V an Lange velde, Ramon J. Havens, Adrie T. A. Zegers-van Duijnhoven, and Vincent C. V enezia. Noise modeling for rf cm os circuit simulation. IEEE Trans. Electron Devices, 50(3):618–632, 2003

    work page 2003

  7. [7]

    Allen, Linda J

    Edward J. Allen, Linda J. S. Allen, Armando Arciniega, an d Priscilla E. Greenwood. Construction of equivalent stochastic differential equation models. Stoch. Anal. Appl., 26(2):274–297, 2008

    work page 2008

  8. [8]

    An introduction to numerical methods fo r stochastic differential equations

    Eckhard Platen. An introduction to numerical methods fo r stochastic differential equations. Acta Numer., 8:197– 246, 1999

    work page 1999

Show all 43 references
  1. [9]

    Desmond J. Higham. An algorithmic introduction to numer ical simulation of stochastic differential equations. SIAM Rev., 43(3):525–546, 2001

    2001

  2. [10]

    Kloeden and Eckhard Platen

    Peter E. Kloeden and Eckhard Platen. Numerical Solution of Stochastic Differential Equations. Stochastic Modelling and Applied Probability. Springer: Berlin, 2011

    2011

  3. [11]

    Pretty, J

    Béla Paláncz, Kent Stewart, József Homlok, Christophe r G. Pretty, J. Geoffrey Chase, and Balázs Benyó. Stochas- tic simulation and parameter estimation of the icing model. IFAC-PapersOnLine, 49(5):218–223, 2016

    2016

  4. [12]

    Akman, James C

    Ozgur E. Akman, James C. W . Locke, Sanyi Tang, Isabelle C arré, Andrew J. Millar, and David A. Rand. Isoform switching facilitates period control in the Neurospora cra ssa circadian clock. Mol. Syst. Biol., 4:164, 2008

    2008

  5. [13]

    Akman, David A

    Ozgur E. Akman, David A. Rand, Paul E. Brown, and Andrew J . Millar. Robustness from flexibility in the fungal circadian clock. BMC Syst. Biol., 4(1):88, 2010

    2010

  6. [14]

    Parameter est imation and model selection in computational biology

    Gabriele Lillacci and Mustafa Khammash. Parameter est imation and model selection in computational biology. PLoS Comput. Biol., 6(3):e1000696, 2010

    2010

  7. [15]

    Akman, Steven Watterson, Andrew Parton, Nigel Binns, Andrew J

    Ozgur E. Akman, Steven Watterson, Andrew Parton, Nigel Binns, Andrew J. Millar, and Peter Ghazal. Digital clocks: simple Boolean models can quantitatively describe circadian systems. J. Roy. Soc. Interface, 9(74):2365– 2382, 2012. 14 SODECL: An Open Source Library for Calculat...

    2012

  8. [16]

    Eleftherios Avramidis and Ozgur E. Akman. Optimisatio n of an exemplar oculomotor model using multi- objective genetic algorithms executed on a gpu-cpu combina tion. BMC Syst. Biol., 11:1, 2017

    2017

  9. [17]

    Akman, and Jon athan E

    Kevin Doherty, Khulood Alyahya, Ozgur E. Akman, and Jon athan E. Fieldsend. Optimisation and landscape analysis of computational biology models: A case study. In Proc. GECCO ’17, 2017, pages 1644–1651, 2017

    2017

  10. [18]

    Cullen, Claudio G

    Kathleen E. Cullen, Claudio G. Rey, Daniel Guitton, and Henrietta L. Galiana. The use of system identification techniques in the analysis of oculomotor burst neuron spike train dynamics. J. Comput. Neurosci., 3(4):347–368, 1996

    1996

  11. [19]

    Johnson and Kristian S

    Jerald B. Johnson and Kristian S. Omland. Model selecti on in ecology and evolution. Trends Ecol. Evol., 19(2):101–108, 2004

    2004

  12. [20]

    Ka andorp, and Joke G

    Maksat Ashyraliyev, Yves Fomekong-Nanfack, Jaap A. Ka andorp, and Joke G. Blom. Systems biology: param- eter estimation for biochemical models. FEBS J., 276(4):886–902, 2009

    2009

  13. [21]

    Systems biology: mode l based evaluation and comparison of potential explanations for given biological data

    Gunnar Cedersund and Jacob Roll. Systems biology: mode l based evaluation and comparison of potential explanations for given biological data. FEBS J., 276(4):903–922, 2009

    2009

  14. [22]

    Diego. F. Slezak, Cecilia Suárez, Guillermo A. Cecchi, Guillermo Marshall, and Gustavo Stolovitzky. When the optimal is not the best: parameter estimation in complex bio logical models. PloS One, 5(10):e13283, 2010

    2010

  15. [23]

    Garibaldi, and Charlie Hodgm an

    Jianyong Sun, Jonathan M. Garibaldi, and Charlie Hodgm an. Parameter estimation using metaheuristics in systems biology: a comprehensive review. IEEE/ACM Trans. Comput. Biol. Bioinform., 9(1):185–202, 2012

    2012

  16. [24]

    Drift analysis and average time compl exity of evolutionary algorithms

    Jun He and Xin Y ao. Drift analysis and average time compl exity of evolutionary algorithms. Artif. Intell., 127(1):57–85, 2001

    2001

  17. [25]

    Population size versus runtime of a simpl e evolutionary algorithm

    Carsten Witt. Population size versus runtime of a simpl e evolutionary algorithm. Theor. Comput. Sci., 403(1):104–120, 2008

    2008

  18. [26]

    Optimiza- tion in biology parameter estimation and the associated opt imization problem

    Gunnar Cedersund, Oscar Samuelsson, Gordon Ball, Jesp er Tegnér, and David Gomez-Cabrero. Optimiza- tion in biology parameter estimation and the associated opt imization problem. In Liesbert Geris and David Gomez-Cabrero, editors, Uncertainty in Biology: A Computational Modeli...

    2016

  19. [27]

    Fieldsend

    Jonathan E. Fieldsend. Elite accumulative sampling st rategies for noisy multi-objective optimisation. In Proc. EMO 2015, 2015, pages 172–186, 2015

    2015

  20. [28]

    Navarro, Nancy Hitschfeld-Kahler, and Lu is Mateu

    Cristóbal A. Navarro, Nancy Hitschfeld-Kahler, and Lu is Mateu. A survey on parallel computing and its appli- cations in data-parallel problems using gpu architectures . Commun. Comput. Phys., 15(2):285–329, 2014

    2014

  21. [29]

    A new era in scientific computing: Domain decomposition methods in hybrid cpu–gpu architectu res

    Manolis Papadrakakis, George Stavroulakis, and Alexa nder Karatarakis. A new era in scientific computing: Domain decomposition methods in hybrid cpu–gpu architectu res. Comput. Methods Appl. Mech. Engrg., 200(13):1490–1508, 2011

    2011

  22. [30]

    De la Asunción and Manuel J

    Marc Mantas, José M. De la Asunción and Manuel J. Castro. An introduction to gpu computing for numerical simulation. In Inmaculada Higueras, Teo Roldán, and Juan J. Torrens, editors, Numerical Simulation in Physics and Engineering, pages 219–251. Springer, 2016

    2016

  23. [31]

    Programming cuda and opencl: A case study using modern c++ libraries

    Denis Demidov, Karsten Ahnert, Karl Rupp, and Peter Got tschling. Programming cuda and opencl: A case study using modern c++ libraries. SIAM J. Sci. Comput., 35(5):C453–C472, 2013

    2013

  24. [32]

    The opencl specification , version 2.1, document revision: 23

    Khronos OpenCL Working Group. The opencl specification , version 2.1, document revision: 23. Technical report, Lee Howes (Ed.), 2015

    2015

  25. [33]

    Stability analys is of numerical schemes for stochastic differential equa- tions

    Y oshihiro Saito and Taketomo Mitsui. Stability analys is of numerical schemes for stochastic differential equa- tions. SIAM J. Numer. Anal., 33(6):2254–2267, 1996

    1996

  26. [34]

    Optimisation and computational methods to model the oculomotor system with focus on nystagmus

    Eleftherios Avramidis. Optimisation and computational methods to model the oculomotor system with focus on nystagmus. PhD thesis, University of Exeter, 2015

    2015

  27. [35]

    Salmon, Mark A

    John K. Salmon, Mark A. Moraes, Ron O. Dror, and David E. S haw. Parallel random numbers: as easy as 1, 2, 3. In SC ’11: Proceedings of 2011 International Conference for High Performance Computing, Networking, Storage and Analysis, 2011, pages 1–12, 2011

    2011

  28. [36]

    G. E. P . Box and Mervin E. Muller. A note on the generation of random normal deviates. Ann. Math. Statist., 29(2):610–611, 1958

    1958

  29. [37]

    Order Function and Macroscopic Mutual E ntrainment in Uniformly Coupled Limit-Cycle Oscil- lators

    Hiroaki Daido. Order Function and Macroscopic Mutual E ntrainment in Uniformly Coupled Limit-Cycle Oscil- lators. Prog. Theor. Phys., 88(6):1213–1218, 1992

    1992

  30. [38]

    Strogatz

    Steven H. Strogatz. From kuramoto to crawford: explori ng the onset of synchronization in populations of coupled oscillators. Physica D, 143(1-4):1–20, 2000. 15 SODECL: An Open Source Library for Calculating Multiple Orb its of a System of Stochastic Differential Equations in ...

    2000

  31. [39]

    Acebrón, Luis L

    Juan A. Acebrón, Luis L. Bonilla, Conrad J. Pérez Vicent e, Félix Ritort, and Renato Spigler. The kuramoto model: A simple paradigm for synchronization phenomena. Rev. Mod. Phys., 77(1):137, 2005

    2005

  32. [40]

    Chaos in symmetric phase oscillator networks

    Christian Bick, Marc Timme, Danilo Paulikat, Dirk Rath lev, and Peter Ashwin. Chaos in symmetric phase oscillator networks. Phys. Rev. Lett., 107(24):244101, 2011

    2011

  33. [41]

    H. Nakao. Phase reduction approach to synchronisation of nonlinear oscillators. Contemp. Phys., 57(2):188–214, 2016

    2016

  34. [42]

    Openmp: an industry st andard api for shared-memory programming

    Leonardo Dagum and Ramesh Menon. Openmp: an industry st andard api for shared-memory programming. IEEE Comput. Sci. Eng., 5(1):46–55, 1998

    1998

  35. [43]

    Akman, Carl Troei n, Andrew J

    Richard Adams, Allan Clark, Azusa Y amaguchi, Neil Hanl on, Nikos Tsorman, Shakir Ali, Galina Lebedeva, Alexey Goltsov, Anatoly Sorokin, Ozgur E. Akman, Carl Troei n, Andrew J. Millar, Igor Goryanin, and Stephen Gilmore. Sbsi: an extensible distributed software infrast ructure...

    2013

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.