A Performance Analysis of Parallel Differential Dynamic Programming on a GPU

Plancher, Brian; Kuindersma, Scott

doi:10.1007/978-3-030-44051-0_38

Cited by 20 publications

(22 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While there are many existing state-of-the-art spatialalgebra-based rigid body dynamics libraries [25], [11], [13], [26], [27], these libraries are not optimized for GPUs [16]. The exception is the recently released NVIDIA Isaac Sim [28] which supports spatial-algebra-based forward simulation, but not gradients, on the GPU.…”

Section: Related Workmentioning

confidence: 99%

“…Despite being highly accurate and optimized, existing implementations of spatial-algebra-based approaches to rigid body dynamics [10] do not take advantage of opportunities for parallelism present in the algorithm, limiting their performance [11], [12], [13], [14]. This is critical because there is natural parallelism in many bottleneck computations involving rigid body dynamics in robotics [15], [16], [9]. For example, the gradient of forward dynamics accounts for 30% to 90% of the total computational time of typical nonlinear model-predictive control (MPC) implementations [17], [18], [12], [16], and is naturally parallel across the discrete points in the trajectory.…”

Section: Introductionmentioning

confidence: 99%

“…This is critical because there is natural parallelism in many bottleneck computations involving rigid body dynamics in robotics [15], [16], [9]. For example, the gradient of forward dynamics accounts for 30% to 90% of the total computational time of typical nonlinear model-predictive control (MPC) implementations [17], [18], [12], [16], and is naturally parallel across the discrete points in the trajectory.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

GRiD: GPU-Accelerated Rigid Body Dynamics with Analytical Gradients

Plancher¹,

Neuman²,

Ghosal³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

We introduce GRiD: a GPU-accelerated library for computing rigid body dynamics with analytical gradients. GRiD was designed to accelerate the nonlinear trajectory optimization subproblem used in state-of-the-art robotic planning, control, and machine learning. Each iteration of nonlinear trajectory optimization requires tens to hundreds of naturally parallel computations of rigid body dynamics and their gradients. GRiD leverages URDF parsing and code generation to deliver optimized dynamics kernels that not only expose GPUfriendly computational patterns, but also take advantage of both fine-grained parallelism within each computation and coarsegrained parallelism between computations. Through this approach, when performing multiple computations of rigid body dynamics algorithms, GRiD provides as much as a 7.6x speedup over a state-of-the-art, multi-threaded CPU implementation, and maintains as much as a 2.6x speedup when accounting for I/O overhead. We release GRiD as an open-source library, so that it can be leveraged by the robotics community to easily and efficiently accelerate rigid body dynamics on the GPU.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

GRiD: GPU-Accelerated Rigid Body Dynamics with Analytical Gradients

Plancher¹,

Neuman²,

Ghosal³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Recent work has focused on making the underlying planning and control algorithms faster. For example, Giftthaler et al [12] introduced a multiple-shooting variant of the trajectory optimizer-iterative linear quadratic regulator [24] which has shown impressive results for real-time nonlinear optimal control of complex robotic systems [28,31].…”

Section: Related Workmentioning

confidence: 99%

Parareal with a learned coarse model for robotic manipulation

Agboh

Grainger

Ruprecht

et al. 2020

Comput. Visual Sci.

View full text Add to dashboard Cite

A key component of many robotics model-based planning and control algorithms is physics predictions, that is, forecasting a sequence of states given an initial state and a sequence of controls. This process is slow and a major computational bottleneck for robotics planning algorithms. Parallel-in-time integration methods can help to leverage parallel computing to accelerate physics predictions and thus planning. The Parareal algorithm iterates between a coarse serial integrator and a fine parallel integrator. A key challenge is to devise a coarse model that is computationally cheap but accurate enough for Parareal to converge quickly. Here, we investigate the use of a deep neural network physics model as a coarse model for Parareal in the context of robotic manipulation. In simulated experiments using the physics engine Mujoco as fine propagator we show that the learned coarse model leads to faster Parareal convergence than a coarse physics-based model. We further show that the learned coarse model allows to apply Parareal to scenarios with multiple objects, where the physics-based coarse model is not applicable. Finally, we conduct experiments on a real robot and show that Parareal predictions are close to real-world physics predictions for robotic pushing of multiple objects. Code (https://doi.org/10.5281/zenodo.3779085) and videos (https://youtu.be/wCh2o1rf-gA) are publicly available.

show abstract

“…In multi-GPUs parallelization for the Image Processing, MCM is realized with the help of two Kepler architectures and remarkable results are achieved [27]. Specifically for the robotics field, various algorithms belonging to the Differential Dynamic Programming (DDP) are parallelized and evaluated on the GPU and significant improvements are recorded in comparison with the multithreaded CPU [28]. The Pipelining approach is an experiment in the GPU implementation of parallel DP [29].…”

Section: Literature Reviewmentioning

confidence: 99%

A Parallelization of Non-Serial Polyadic Dynamic Programming on GPU

Diwan

Tembhurne

2019

CIT. J. comput. inf. technol

View full text Add to dashboard Cite

Parallelization of Non-Serial Polyadic Dynamic Programming (NPDP) on high-throughput manycore architectures, such as NVIDIA GPUs, suffers from load imbalance, i.e. non-optimal mapping between the sub-problems of NPDP and the processing elements of the GPU. NPDP exhibits non-uniformity in the number of subproblems as well as computational complexity across the phases. In NPDP parallelization, phases are computed sequentially whereas subproblems of each phase are computed concurrently. Therefore, it is essential to effectively map the subproblems of each phase to the processing elements while implementing thread level parallelism. We propose an adaptive Generalized Mapping Method (GMM) for NPDP parallelization that utilizes the GPU for efficient mapping of subproblems onto processing threads in each phase. Input-size and targeted GPU decide the computing power and the best mapping for each phase in NPDP parallelization. The performance of GMM is compared with different conventional parallelization approaches. For sufficiently large inputs, our technique outperforms the state-of-the-art conventional parallelization approach and achieves a significant speedup of a factor 30. We also summarize the general heuristics for achieving better gain in the NPDP parallelization.

show abstract

A Performance Analysis of Parallel Differential Dynamic Programming on a GPU

Cited by 20 publications

References 25 publications

GRiD: GPU-Accelerated Rigid Body Dynamics with Analytical Gradients

GRiD: GPU-Accelerated Rigid Body Dynamics with Analytical Gradients

Parareal with a learned coarse model for robotic manipulation

A Parallelization of Non-Serial Polyadic Dynamic Programming on GPU

Contact Info

Product

Resources

About