pith. sign in

arxiv: 2605.15875 · v1 · pith:7F5CF66Nnew · submitted 2026-05-15 · 💻 cs.GR

Distributed Affine Body Dynamics with Adaptive Consensus

Jiafeng Liu , Wenhui Zhou , Xinming Pei , Yifan Peng , Huamin Wang , Yin Yang , Lei Lan , Weiwei Xu This is my paper

Pith reviewed 2026-05-19 18:24 UTC · model grok-4.3

classification 💻 cs.GR
keywords distributed simulationaffine body dynamicsincremental potential contactconsensus ADMMparallel computingrigid body dynamicscontact simulationGPU clusters
0
0 comments X

The pith

A consensus-based ADMM scheme distributes Affine Body Dynamics simulations across multiple compute nodes while preserving IPC non-penetration guarantees.

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

Affine Body Dynamics inside the Incremental Potential Contact framework simulates extremely stiff solids with near-rigid behavior and strict non-penetration. The global barrier constraints in IPC prevent easy scaling to multiple GPUs or compute nodes. The paper breaks the problem into local ABD subproblems solved independently in parallel on each node. A global consensus step then aligns results on shared boundary bodies. This would matter to a reader because it removes the single-machine limit on scene size while claiming to retain the original robustness and consistency.

Core claim

The authors propose a distributed formulation of ABD using a consensus-based ADMM scheme. Each compute node solves its local ABD subproblem in parallel, followed by a global consensus step that enforces consistency among shared boundary bodies. The method preserves IPC-level robustness and global consistency under distributed execution.

What carries the argument

consensus-based ADMM scheme that reconciles local ABD subproblems across nodes via shared boundary bodies

Load-bearing premise

Local ABD subproblems can be solved independently on each node and then reconciled through a global consensus step without introducing penetrations or losing the strict non-penetration guarantees of the original IPC framework.

What would settle it

Running an identical scene on a single node versus the distributed system and finding a penetration between bodies on separate nodes that does not appear in the single-node run would disprove preservation of non-penetration.

Figures

Figures reproduced from arXiv: 2605.15875 by Huamin Wang, Jiafeng Liu, Lei Lan, Weiwei Xu, Wenhui Zhou, Xinming Pei, Yifan Peng, Yin Yang.

Figure 1
Figure 1. Figure 1: A contact-rich simulation in which 5K “Pokémon” ABD bodies (6M triangles in total) fall into a narrow funnel under a two-worker distributed simulation [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Partition model. The entire scene is partitioned into two regions [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Illustration of our CCD check process. We show one shared body [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: A CCD-check corner case with a fixed blocking obstacle. In both rows, [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 6
Figure 6. Figure 6: Scaling results on the same scene partitioned into [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Falling foods. A falling and piling scene with food-shaped rigid bodies, containing over 16𝑀 triangles and reaching a peak of 164M collision candidates and 222𝐾 active contact pairs. Using six workers, the distributed solver achieves a 5.2× speedup over a single-worker run. Even under such dense contact interactions, our system remains globally penetration-free and continues to exhibit near-linear scaling … view at source ↗
Figure 8
Figure 8. Figure 8: Industrial parts and robotic arm. A large-scale scenario with 7𝐾 industrial parts and a robotic arm, totaling over 13𝑀 triangles and hundreds of millions of contact candidates. The scene cannot be executed on a single machine due to memory constraints. By distributing geometry and contact processing across workers, our system significantly reduces per-worker memory footprint, enabling simulation at scales … view at source ↗
Figure 9
Figure 9. Figure 9: Rotating drum with shells. Dense-contact simulation of 2𝐾 complex-shaped rigid bodies(e.g., starfish, conches, and shells) continuously agitated in a rotating drum, generating up to 54𝑀 collision candidates. The two-worker distributed solver achieves a 1.5× speedup over a single-machine run. Even in this challenging setting with thin shells and sharp features undergoing sustained agitation, our solver main… view at source ↗
Figure 10
Figure 10. Figure 10: Solid-fluid coupling. Alongside distributed ABD, our system also supports distributed fluid simulation and fluid–solid coupling. In this demo, four workers run distributed IISPH with 1M fluid particles and an ABD solid scene of 7 bodies totaling 220K triangles, with impulse-based two-way solid–fluid coupling. These results show that our multi-node architecture can support coupled multiphysics within a uni… view at source ↗
Figure 11
Figure 11. Figure 11: Mass scaling for penalty initialization. We vary density to scale [PITH_FULL_IMAGE:figures/full_fig_p011_11.png] view at source ↗
Figure 13
Figure 13. Figure 13: Experimental scene for evaluating the effect of mass and [PITH_FULL_IMAGE:figures/full_fig_p011_13.png] view at source ↗
read the original abstract

Affine Body Dynamics (ABD) within the Incremental Potential Contact (IPC) framework provides accurate simulation of extremely stiff solids exhibiting near-rigid behavior, with strict non-penetration guarantees. However, IPC's globally coupled barrier constraints hinder scalable execution across multiple GPUs and compute nodes. We propose a distributed formulation of ABD using a consensus-based ADMM scheme. Each compute node solves its local ABD subproblem in parallel, followed by a global consensus step that enforces consistency among shared boundary bodies. The proposed method preserves IPC-level robustness and global consistency under distributed execution. Experiments demonstrate stable convergence, non-penetration, and efficient scaling on large-scale scenes across multiple nodes.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript presents a distributed formulation of Affine Body Dynamics (ABD) integrated with the Incremental Potential Contact (IPC) framework. It uses a consensus-based ADMM scheme in which each compute node independently solves a local ABD subproblem, followed by a global consensus step that enforces consistency on the positions and velocities of shared boundary bodies. The central claim is that this distributed approach preserves IPC-level robustness, strict non-penetration guarantees, and global consistency while enabling scalable execution across multiple GPUs and compute nodes. Experiments on large-scale scenes are cited to show stable convergence and efficient scaling.

Significance. If the non-penetration guarantees hold for all contacts (including those spanning node boundaries), the work would enable practical large-scale simulation of stiff, near-rigid solids on distributed hardware without sacrificing the robustness properties of centralized IPC. This would be a useful contribution to scalable physics simulation in computer graphics. The application of ADMM consensus to ABD is a reasonable technical choice, but its ability to replicate the original barrier constraints is the key determinant of impact.

major comments (2)
  1. [Method / consensus step] Method description (distributed formulation and consensus step): the central claim requires that local ABD subproblems plus ADMM consensus on shared boundary bodies yields the same strict non-penetration as centralized IPC. In a partitioned domain, contacts whose candidate pairs lie across node boundaries are invisible to any single local subproblem. The consensus step is described as operating only on the positions/velocities of the shared bodies; unless the barrier potential for every cross-boundary contact is explicitly added to the consensus objective (or the contact set is globally synchronized before the ADMM loop), the distributed solve can converge to a point that violates the IPC barrier while satisfying only the local and consensus terms. No derivation, error analysis, or pseudocode is supplied to show that this synchronization occurs.
  2. [Abstract and Experiments] Abstract and § on experiments: the claim of 'IPC-level robustness and non-penetration' is asserted without supporting derivation details, convergence analysis for the distributed barrier terms, or quantitative checks (e.g., maximum penetration depth or barrier energy values) on cross-node contact pairs in the reported large-scale scenes.
minor comments (2)
  1. [Title and Method] The title refers to 'Adaptive Consensus' but the method sketch describes a standard ADMM consensus step; a brief clarification of the adaptive mechanism (if any) would improve readability.
  2. [Notation] Notation for the local subproblem objectives and the global consensus objective should be aligned more explicitly with the original IPC formulation to make the preservation argument easier to follow.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their thorough review and valuable feedback on our manuscript. We address each of the major comments below and outline the revisions we will make to improve the clarity and completeness of the presentation.

read point-by-point responses

  1. Referee: [Method / consensus step] Method description (distributed formulation and consensus step): the central claim requires that local ABD subproblems plus ADMM consensus on shared boundary bodies yields the same strict non-penetration as centralized IPC. In a partitioned domain, contacts whose candidate pairs lie across node boundaries are invisible to any single local subproblem. The consensus step is described as operating only on the positions/velocities of the shared bodies; unless the barrier potential for every cross-boundary contact is explicitly added to the consensus objective (or the contact set is globally synchronized before the ADMM loop), the distributed solve can converge to a point that violates the IPC barrier while satisfying only the local and consensus terms. No derivation, error analysis, or pseudocode is supplied to show that this synchronization occurs.

    Authors: We appreciate the referee highlighting this critical point regarding cross-boundary contacts. In our formulation, global contact detection identifies all candidate pairs prior to partitioning, and cross-boundary pairs contribute their barrier potentials directly to the consensus objective within the ADMM loop. The augmented Lagrangian therefore incorporates these terms, ensuring the distributed solve respects the full IPC barrier set. We will add an explicit derivation of the distributed barrier inclusion, a brief error analysis showing equivalence to the centralized case under consensus convergence, and pseudocode for the contact synchronization step in the revised manuscript. revision: yes

  2. Referee: [Abstract and Experiments] Abstract and § on experiments: the claim of 'IPC-level robustness and non-penetration' is asserted without supporting derivation details, convergence analysis for the distributed barrier terms, or quantitative checks (e.g., maximum penetration depth or barrier energy values) on cross-node contact pairs in the reported large-scale scenes.

    Authors: We agree that the current presentation would benefit from stronger empirical support for the non-penetration claims on cross-node contacts. In the revised experiments section we will report maximum penetration depths and barrier energy values measured specifically on cross-node contact pairs, together with convergence plots of the distributed barrier residuals. These additions will be placed alongside the existing scaling results. revision: yes

Circularity Check

0 steps flagged

No circularity detected in derivation chain

full rationale

The paper presents a new distributed formulation of ABD via consensus ADMM, with each node solving independent local subproblems followed by a global consensus step on boundary bodies. This is framed as an extension that preserves IPC non-penetration guarantees rather than deriving any core quantity by re-fitting or redefining prior results. No equations or steps reduce a prediction to a fitted input by construction, and no load-bearing claim relies on a self-citation chain that itself assumes the target result. The method is self-contained against the centralized IPC baseline, with the distributed reconciliation treated as an independent algorithmic contribution.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the convergence of the ADMM consensus step and the assumption that problem partitioning preserves global non-penetration; no free parameters or invented entities are mentioned in the abstract.

axioms (1)
  • domain assumption ADMM consensus step enforces global consistency among shared boundary bodies without violating non-penetration
    Invoked to guarantee that distributed execution matches the original IPC properties.

pith-pipeline@v0.9.0 · 5645 in / 1000 out tokens · 33132 ms · 2026-05-19T18:24:38.846089+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

76 extracted references · 76 canonical work pages · 1 internal anchor

  1. [1]

    Interactive augmented reality storytelling guided by scene semantics,

    Lan, Lei and Kaufman, Danny M. and Li, Minchen and Jiang, Chenfanfu and Yang, Yin , title =. 2022 , issue_date =. doi:10.1145/3528223.3530064 , month = jul, articleno =

    work page doi:10.1145/3528223.3530064 2022

  2. [2]

    ACM Transactions on Graphics (TOG) , volume=

    A unified newton barrier method for multibody dynamics , author=. ACM Transactions on Graphics (TOG) , volume=. 2022 , publisher=

    work page 2022

  3. [3]

    arXiv preprint arXiv:2411.06224 , year=

    StiffGIPC: Advancing GPU IPC for stiff affine-deformable simulation , author=. arXiv preprint arXiv:2411.06224 , year=

    work page arXiv

  4. [4]

    Kaufman , title =

    Minchen Li and Zachary Ferguson and Teseo Schneider and Timothy Langlois and Denis Zorin and Daniele Panozzo and Chenfanfu Jiang and Danny M. Kaufman , title =. ACM Trans. Graph. (SIGGRAPH) , year =

    work page

  5. [5]

    2023 , publisher=

    Distributed Simulation of Large Multi-body Systems , author=. 2023 , publisher=

    work page 2023

  6. [6]

    Seminal Graphics Papers: Pushing the Boundaries, Volume 2 , pages=

    Projective dynamics: Fusing constraint projections for fast simulation , author=. Seminal Graphics Papers: Pushing the Boundaries, Volume 2 , pages=

    work page

  7. [7]

    and Li, Jie and Narain, Rahul , journal=

    Overby, Matthew and Brown, George E. and Li, Jie and Narain, Rahul , journal=. ADMM Projective Dynamics: Fast Simulation of Hyperelastic Models with Dynamic Constraints , year=

    work page

  8. [8]

    Foundations and Trends

    Distributed optimization and statistical learning via the alternating direction method of multipliers , author=. Foundations and Trends. 2011 , publisher=

    work page 2011

  9. [9]

    Computer Graphics Forum , volume=

    Fast BVH construction on GPUs , author=. Computer Graphics Forum , volume=. 2009 , organization=

    work page 2009

  10. [10]

    ACM SIGGRAPH 2023 Conference Proceedings , pages=

    Interactive hair simulation on the GPU using ADMM , author=. ACM SIGGRAPH 2023 Conference Proceedings , pages=

    work page 2023

  11. [11]

    ACM Transactions on Graphics (TOG) , volume=

    Simple and scalable frictional contacts for thin nodal objects , author=. ACM Transactions on Graphics (TOG) , volume=. 2020 , publisher=

    work page 2020

  12. [12]

    ACM Transactions on Graphics (TOG) , volume=

    Wrapd: weighted rotation-aware admm for parameterization and deformation , author=. ACM Transactions on Graphics (TOG) , volume=. 2021 , publisher=

    work page 2021

  13. [13]

    ACM Transactions on Graphics (TOG) , volume=

    Silly rubber: an implicit material point method for simulating non-equilibrated viscoelastic and elastoplastic solids , author=. ACM Transactions on Graphics (TOG) , volume=. 2019 , publisher=

    work page 2019

  14. [14]

    ACM Transactions on Graphics (TOG) , volume=

    Penetration-free projective dynamics on the GPU , author=. ACM Transactions on Graphics (TOG) , volume=. 2022 , publisher=

    work page 2022

  15. [15]

    ACM Transactions on Graphics , volume=

    Fast GPU-based two-way continuous collision handling , author=. ACM Transactions on Graphics , volume=. 2023 , publisher=

    work page 2023

  16. [16]

    IEEE Transactions on Signal Processing , volume=

    Multi-agent distributed optimization via inexact consensus ADMM , author=. IEEE Transactions on Signal Processing , volume=. 2014 , publisher=

    work page 2014

  17. [17]

    Computational Optimization and Applications , volume=

    An inexact ADMM for separable nonconvex and nonsmooth optimization , author=. Computational Optimization and Applications , volume=. 2025 , publisher=

    work page 2025

  18. [18]

    ACM Transactions on Graphics (TOG) , volume=

    A massively parallel and scalable multi-GPU material point method , author=. ACM Transactions on Graphics (TOG) , volume=. 2020 , publisher=

    work page 2020

  19. [19]

    ACM Transactions on Graphics , volume=

    A sparse distributed gigascale resolution material point method , author=. ACM Transactions on Graphics , volume=. 2023 , publisher=

    work page 2023

  20. [20]

    ACM Transactions on Graphics (TOG) , volume=

    Accelerating ADMM for efficient simulation and optimization , author=. ACM Transactions on Graphics (TOG) , volume=. 2019 , publisher=

    work page 2019

  21. [21]

    ACM Transactions on Graphics , volume=

    Anderson Acceleration for Geometry Optimization and Physics Simulation , author=. ACM Transactions on Graphics , volume=

    work page

  22. [22]

    SIAM Journal on Scientific Computing , volume=

    Anderson Accelerated Douglas--Rachford Splitting , author=. SIAM Journal on Scientific Computing , volume=. 2020 , publisher=

    work page 2020

  23. [23]

    Computer Graphics Forum , volume=

    Anderson acceleration for nonconvex ADMM based on Douglas-Rachford splitting , author=. Computer Graphics Forum , volume=. 2020 , organization=

    work page 2020

  24. [24]

    ACM Transactions on Graphics (TOG) , volume=

    Scanerf: Scalable bundle-adjusting neural radiance fields for large-scale scene rendering , author=. ACM Transactions on Graphics (TOG) , volume=. 2023 , publisher=

    work page 2023

  25. [25]

    IEEE Access , volume=

    Distributed inexact consensus-based ADMM method for multi-agent unconstrained optimization problem , author=. IEEE Access , volume=. 2019 , publisher=

    work page 2019

  26. [26]

    IEEE Transactions on Parallel and Distributed Systems , volume=

    Advances in multi-GPU smoothed particle hydrodynamics simulations , author=. IEEE Transactions on Parallel and Distributed Systems , volume=. 2012 , publisher=

    work page 2012

  27. [27]

    IEEE transactions on visualization and computer graphics , volume=

    Implicit incompressible SPH , author=. IEEE transactions on visualization and computer graphics , volume=. 2013 , publisher=

    work page 2013

  28. [28]

    ACM Transactions on Graphics (TOG) , volume=

    Automatic quantization for physics-based simulation , author=. ACM Transactions on Graphics (TOG) , volume=. 2022 , publisher=

    work page 2022

  29. [29]

    Computer Graphics Forum , volume=

    Primal-dual optimization for fluids , author=. Computer Graphics Forum , volume=. 2017 , organization=

    work page 2017

  30. [30]

    ACM Transactions on Graphics (ToG) , volume=

    Efficient solver for spacetime control of smoke , author=. ACM Transactions on Graphics (ToG) , volume=. 2017 , publisher=

    work page 2017

  31. [31]

    Codimensional incremental potential contact

    Codimensional incremental potential contact , author=. arXiv preprint arXiv:2012.04457 , year=

    work page arXiv 2012

  32. [32]

    arXiv preprint arXiv:2403.19272 , year=

    Efficient GPU cloth simulation with non-distance barriers and subspace reuse , author=. arXiv preprint arXiv:2403.19272 , year=

    work page arXiv

  33. [33]

    SIGGRAPH Asia 2023 Conference Papers , pages=

    Subspace-preconditioned gpu projective dynamics with contact for cloth simulation , author=. SIGGRAPH Asia 2023 Conference Papers , pages=

    work page 2023

  34. [34]

    ACM Transactions on Graphics , volume=

    Intersection-free rigid body dynamics , author=. ACM Transactions on Graphics , volume=

    work page

  35. [35]

    ACM Transactions on Graphics (TOG) , volume=

    A contact proxy splitting method for Lagrangian solid-fluid coupling , author=. ACM Transactions on Graphics (TOG) , volume=. 2023 , publisher=

    work page 2023

  36. [36]

    2023 , issue_date =

    Lan, Lei and Li, Minchen and Jiang, Chenfanfu and Wang, Huamin and Yang, Yin , title =. 2023 , issue_date =. doi:10.1145/3592104 , journal =

    work page doi:10.1145/3592104 2023

  37. [37]

    ACM Transactions on Graphics , volume=

    GIPC: Fast and stable gauss-newton optimization of IPC barrier energy , author=. ACM Transactions on Graphics , volume=. 2024 , publisher=

    work page 2024

  38. [38]

    ACM Transactions on Graphics (TOG) , volume=

    A GPU-based multilevel additive schwarz preconditioner for cloth and deformable body simulation , author=. ACM Transactions on Graphics (TOG) , volume=. 2022 , publisher=

    work page 2022

  39. [39]

    ACM Transactions on Graphics (TOG) , volume=

    Barrier-Augmented Lagrangian for GPU-based Elastodynamic Contact , author=. ACM Transactions on Graphics (TOG) , volume=. 2024 , publisher=

    work page 2024

  40. [40]

    ACM Transactions on Graphics (TOG) , volume=

    I-Cloth: Incremental collision handling for GPU-based interactive cloth simulation , author=. ACM Transactions on Graphics (TOG) , volume=. 2018 , publisher=

    work page 2018

  41. [41]

    ACM Transactions on Graphics (TOG) , volume=

    P-cloth: interactive complex cloth simulation on multi-GPU systems using dynamic matrix assembly and pipelined implicit integrators , author=. ACM Transactions on Graphics (TOG) , volume=. 2020 , publisher=

    work page 2020

  42. [42]

    Computer Graphics Forum , volume=

    Accelerating Distributed Graphical Fluid Simulations with Micro-partitioning , author=. Computer Graphics Forum , volume=. 2020 , organization=

    work page 2020

  43. [43]

    SC'12: Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis , pages=

    Legion: Expressing locality and independence with logical regions , author=. SC'12: Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis , pages=. 2012 , organization=

    work page 2012

  44. [44]

    ACM Transactions on Graphics (TOG) , volume=

    A scalable schur-complement fluids solver for heterogeneous compute platforms , author=. ACM Transactions on Graphics (TOG) , volume=. 2016 , publisher=

    work page 2016

  45. [45]

    Proceedings of the thirteenth euroSys conference , pages=

    Decoupling the control plane from program control flow for flexibility and performance in cloud computing , author=. Proceedings of the thirteenth euroSys conference , pages=

    work page

  46. [46]

    ACM Transactions on Graphics (TOG) , volume=

    Automatically distributing eulerian and hybrid fluid simulations in the cloud , author=. ACM Transactions on Graphics (TOG) , volume=. 2018 , publisher=

    work page 2018

  47. [47]

    2017 USENIX Annual Technical Conference (USENIX ATC 17) , pages=

    Execution templates: Caching control plane decisions for strong scaling of data analytics , author=. 2017 USENIX Annual Technical Conference (USENIX ATC 17) , pages=

    work page 2017

  48. [48]

    ACM Transactions on Graphics (TOG) , volume=

    Loki: a unified multiphysics simulation framework for production , author=. ACM Transactions on Graphics (TOG) , volume=. 2022 , publisher=

    work page 2022

  49. [49]

    2020 , school=

    Distributed real-time physics for scalable and streamed games and simulation , author=. 2020 , school=

    work page 2020

  50. [50]

    A survey of ADMM variants for distributed optimization: Problems, algorithms and features,

    A survey of ADMM variants for distributed optimization: Problems, algorithms and features , author=. arXiv preprint arXiv:2208.03700 , year=

    work page arXiv

  51. [51]

    The Visual Computer , volume=

    Efficient frictional contacts for soft body dynamics via ADMM , author=. The Visual Computer , volume=. 2024 , publisher=

    work page 2024

  52. [52]

    A Survey: Learning Embodied Intelligence from Physical Simulators and World Models , doi =

    Long, Xiaoxiao and Zhao, Qingrui and Zhang, Kaiwen and Zhang, Zihao and Wang, Dingrui and Liu, Yumeng and Shu, Zhengjie and Lu, Yi and Wang, Shouzheng and Wei, Xinzhe and Li, Wei and Yin, Wei and Yao, Yao and Pan, Jia and Shen, Qiu and Yang, Ruigang and Cao, Xun and Dai, Qionghai , year =. A Survey: Learning Embodied Intelligence from Physical Simulators ...

    work page

  53. [53]

    IEEE Transactions on Visualization and Computer Graphics , volume=

    Direct forcing for lagrangian rigid-fluid coupling , author=. IEEE Transactions on Visualization and Computer Graphics , volume=. 2009 , publisher=

    work page 2009

  54. [54]

    Science Robotics , volume=

    Learning contact-rich whole-body manipulation with example-guided reinforcement learning , author=. Science Robotics , volume=. 2025 , publisher=

    work page 2025

  55. [55]

    International Journal for Numerical Methods in Engineering , volume=

    Solving variational inequalities and cone complementarity problems in nonsmooth dynamics using the alternating direction method of multipliers , author=. International Journal for Numerical Methods in Engineering , volume=. 2021 , publisher=

    work page 2021

  56. [56]

    ACM Transactions on Graphics (TOG) , volume=

    Offset Geometric Contact , author=. ACM Transactions on Graphics (TOG) , volume=. 2025 , publisher=

    work page 2025

  57. [57]

    arXiv preprint arXiv:2512.12151 , year=

    Robust and Efficient Penetration-Free Elastodynamics without Barriers , author=. arXiv preprint arXiv:2512.12151 , year=

    work page arXiv

  58. [58]

    Unified treatment of contact, friction and shock-propagation in rigid body animation , year =

    Chen, Yi-Lu and Ly, Micka\". Unified treatment of contact, friction and shock-propagation in rigid body animation , year =. doi:10.1145/3606037.3606836 , articleno =

    work page doi:10.1145/3606037.3606836

  59. [59]

    arXiv preprint arXiv:2302.14344 , year=

    Modular and parallelizable multibody physics simulation via subsystem-based ADMM , author=. arXiv preprint arXiv:2302.14344 , year=

    work page arXiv

  60. [60]

    ICRA 2025 Workshop''Handy Moves: Dexterity in Multi-Fingered Hands''Paper Submission , year=

    GPU-Accelerated Subsystem-Based ADMM for Large-Scale Interactive Simulation , author=. ICRA 2025 Workshop''Handy Moves: Dexterity in Multi-Fingered Hands''Paper Submission , year=

    work page 2025

  61. [61]

    2025 , issue_date =

    Lan, Lei and Lu, Zixuan and Yuan, Chun and Xu, Weiwei and Su, Hao and Wang, Huamin and Jiang, Chenfanfu and Yang, Yin , title =. 2025 , issue_date =. doi:10.1145/3731183 , journal =

    work page doi:10.1145/3731183 2025

  62. [62]

    instrument society of america research triangle park nc , year=

    PID controllers: Theory, Design and Tuning , author=. instrument society of america research triangle park nc , year=

    work page

  63. [63]

    ACM Transactions on Graphics (TOG) , volume=

    Mass splitting for jitter-free parallel rigid body simulation , author=. ACM Transactions on Graphics (TOG) , volume=. 2012 , publisher=

    work page 2012

  64. [64]

    2004 , publisher=

    Real-time collision detection , author=. 2004 , publisher=

    work page 2004

  65. [65]

    ACM Transaction on Graphics (Proceedings of SIGGRAPH Asia) , volume =

    Li, Cheng and Tang, Min and Tong, Ruofeng and Cai, Ming and Zhao, Jieyi and Manocha, Dinesh , title =. ACM Transaction on Graphics (Proceedings of SIGGRAPH Asia) , volume =

    work page

  66. [66]

    IEEE Journal on Selected Areas in Communications , volume=

    A tutorial on decomposition methods for network utility maximization , author=. IEEE Journal on Selected Areas in Communications , volume=. 2006 , publisher=

    work page 2006

  67. [67]

    Acta numerica , volume=

    Domain decomposition algorithms , author=. Acta numerica , volume=. 1994 , publisher=

    work page 1994

  68. [68]

    IEEE Transactions on automatic control , volume=

    Distributed subgradient methods for multi-agent optimization , author=. IEEE Transactions on automatic control , volume=. 2009 , publisher=

    work page 2009

  69. [69]

    SIAM Journal on Optimization , volume=

    Extra: An exact first-order algorithm for decentralized consensus optimization , author=. SIAM Journal on Optimization , volume=. 2015 , publisher=

    work page 2015

  70. [70]

    IEEE Transactions on Signal Processing , volume=

    Exact diffusion for distributed optimization and learning—Part I: Algorithm development , author=. IEEE Transactions on Signal Processing , volume=. 2018 , publisher=

    work page 2018

  71. [71]

    Communications of the ACM , volume=

    MapReduce: simplified data processing on large clusters , author=. Communications of the ACM , volume=. 2008 , publisher=

    work page 2008

  72. [72]

    Gropp , title =

    William D. Gropp , title =. High Performance Computing --- HiPC 2001 , editor =. 2001 , publisher =

    work page 2001

  73. [73]

    DryadLINQ: A system for general-purpose distributed data-parallel computing using a high-level language , author=. Proc. LSDS-IR , volume=

    work page

  74. [74]

    9th USENIX symposium on networked systems design and implementation (NSDI 12) , pages=

    Resilient distributed datasets: A \ Fault-Tolerant \ abstraction for \ In-Memory \ cluster computing , author=. 9th USENIX symposium on networked systems design and implementation (NSDI 12) , pages=

    work page

  75. [75]

    Distributed GraphLab: A Framework for Machine Learning in the Cloud

    Distributed graphlab: A framework for machine learning in the cloud , author=. arXiv preprint arXiv:1204.6078 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  76. [76]

    2011 , publisher=

    Scaling up machine learning: Parallel and distributed approaches , author=. 2011 , publisher=

    work page 2011