CuPIQP: A Batched and Differentiable QP Solver on GPUs for Robot Learning and Control
Overview
This research presents cuPIQP, a GPU-based quadratic programming solver utilizing a proximal interior-point method to support dense, sparse, and multistage problems. Benchmarks show solve-time speedups of up to 18.8x over Moreau and 4.26x over eight-core PIQP for batches of up to 16384 QPs. In end-to-end learning tasks, it accelerates training by 4.6x over QPAX and 7.8x over Moreau, offering efficient infrastructure for integrating model-based optimization into large-scale robotics pipelines.
Generated from attributed reports · Updated 2 hours ago
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
Reported quantity · solve-time speedup: 18.8 other · Basis not reported
“ation decisions. Benchmarks with up to 16384 QPs per batch demonstrate solve-time speedups of up to $18.8\times$ over \textsc{Moreau} and $4.26\times$ over eight-core \textsc{PIQP}. In an end-to-end learning task, \textsc{cuPIQP} accelerates training by $4.6\times$ over \textsc{QPAX} and $7.8\times$ over \textsc{Moreau}. By reducing the computational cost, \textsc{cuPIQP} provides efficient, reusable infrastructure f”
Exact source · revision 1Source owner not reported
Reported quantity · speedup over eight-core PIQP: 4.26 other · Basis not reported
“ation decisions. Benchmarks with up to 16384 QPs per batch demonstrate solve-time speedups of up to $18.8\times$ over \textsc{Moreau} and $4.26\times$ over eight-core \textsc{PIQP}. In an end-to-end learning task, \textsc{cuPIQP} accelerates training by $4.6\times$ over \textsc{QPAX} and $7.8\times$ over \textsc{Moreau}. By reducing the computational cost, \textsc{cuPIQP} provides efficient, reusable infrastructure f”
Exact source · revision 1Source owner not reported
Reported quantity · training acceleration: 4.6 other · Basis not reported
“ation decisions. Benchmarks with up to 16384 QPs per batch demonstrate solve-time speedups of up to $18.8\times$ over \textsc{Moreau} and $4.26\times$ over eight-core \textsc{PIQP}. In an end-to-end learning task, \textsc{cuPIQP} accelerates training by $4.6\times$ over \textsc{QPAX} and $7.8\times$ over \textsc{Moreau}. By reducing the computational cost, \textsc{cuPIQP} provides efficient, reusable infrastructure f”
Exact source · revision 1Source owner not reported
Reported quantity · training acceleration over Moreau: 7.8 other · Basis not reported
“ation decisions. Benchmarks with up to 16384 QPs per batch demonstrate solve-time speedups of up to $18.8\times$ over \textsc{Moreau} and $4.26\times$ over eight-core \textsc{PIQP}. In an end-to-end learning task, \textsc{cuPIQP} accelerates training by $4.6\times$ over \textsc{QPAX} and $7.8\times$ over \textsc{Moreau}. By reducing the computational cost, \textsc{cuPIQP} provides efficient, reusable infrastructure f”
Exact source · revision 1Source owner not reported
Reported quantity · QP batch size: 16384 other · Basis not reported
“ation decisions. Benchmarks with up to 16384 QPs per batch demonstrate solve-time speedups of up to $18.8\times$ over \textsc{Moreau} and $4.26\times$ over eight-core \textsc{PIQP}. In an end-to-end learning task, \textsc{cuPIQP} accelerates training by $4.6\times$ over \textsc{QPAX} and $7.8\times$ over \textsc{Moreau}. By reducing the computational cost, \textsc{cuPIQP} provides efficient, reusable infrastructure f”
Exact source · revision 1Source owner not reported
Report timeline
Follow attributed reports and material updates.
- arXiv Robotics — research abstractsSignalCuPIQP: A Batched and Differentiable QP Solver on GPUs for Robot Learning and Control
This research presents cuPIQP, a GPU-based quadratic programming solver utilizing a proximal interior-point method to support dense, sparse, and multistage problems. Benchmarks show solve-time speedups of up to 18.8x over Moreau and 4.26x over eight-core PIQP for batches of up to 16384 QPs. In end-to-end learning tasks, it accelerates training by 4.6x over QPAX and 7.8x over Moreau, offering efficient infrastructure for integrating model-based optimization into large-scale robotics pipelines.
Event coverage history
There is not enough continuous observation data to show a trend.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).