Skip to content

Event

cuPIQP: A GPU Batched Differentiable QP Solver Released

1 reports · 1 sources

Fenglong Song and other authors have released cuPIQP, a GPU batched differentiable quadratic programming (QP) solver based on the proximal interior-point method, which uniformly supports dense, sparse, and multi-stage problems and targets robot learning and control. The authors report that in benchmarks with up to 16,384 QPs per batch, cuPIQP solves up to 18.8 times faster than Moreau and 4.26 times faster than eight-core PIQP; in end-to-end learning tasks, training is 4.6 times faster than QPAX and 7.8 times faster than Moreau, respectively. All of the above speedup figures come from benchmarks and tasks reported by the authors and have not yet been independently verified.

Report timeline

  1. cuPIQP: A Differentiable Batched QP GPU Solver for Robot Learning and Control

    cuPIQP is a GPU batched differentiable quadratic programming (QP) solver based on a proximal interior-point method, with unified support for dense, sparse, and multistage problems. In benchmarks with up to 16,384 QPs per batch, it solves up to 18.8 times faster than Moreau and 4.26 times faster than eight-core PIQP; in an end-to-end learning task, training is accelerated by 4.6 times over QPAX and 7.8 times over Moreau.

    arXiv cs.RO ·

Related topics

Robot models and learningSoftware and compute systemsResearch papers

Interest trend

Current
3
Peak
4Nov 4 01:00
Past 24 hours
–

The trend compares participants observed throughout the same period, so its range may be smaller than the current interest count. Move or tap to inspect each hour; use the arrow keys on a keyboard.