eprintid: 2779
rev_number: 11
eprint_status: archive
userid: 44
dir: disk0/00/00/27/79
datestamp: 2015-10-22 13:41:13
lastmod: 2015-10-22 13:41:13
status_changed: 2015-10-22 13:41:13
type: conference_item
metadata_visibility: show
creators_name: Sampathirao, Ajay Kumar
creators_name: Sopasakis, Pantelis
creators_name: Bemporad, Alberto
creators_name: Patrinos, Panagiotis
creators_id: 
creators_id: pantelis.sopasakis@imtlucca.it
creators_id: alberto.bemporad@imtlucca.it
creators_id: panagiotis.patrinos@imtlucca.it
title: Distributed solution of stochastic optimal control problems on GPUs
ispublished: pub
subjects: QA
subjects: QA75
subjects: T1
subjects: TK
divisions: CSA
full_text_status: public
pres_type: speech
abstract: Stochastic optimal control problems arise in many applications and are, in principle, large-scale involving up to millions of decision variables. Their applicability in control applications is often limited by the availability of algorithms that can solve them efficiently and within  the sampling time of the controlled system.

In this paper we propose a dual accelerated proximal gradient algorithm which is amenable to parallelization and demonstrate that its GPU implementation affords high speed-up values (with respect to a CPU implementation) and greatly outperforms well-established commercial optimizers such as Gurobi.
date: 2015
date_type: submitted
publisher: IEEE
event_title: 54th IEEE Conference on Decision and Control
event_location: Osaka, Japan
event_dates: 15-18 December, 2015
event_type: conference
refereed: TRUE
book_title: IEEE Conference on Decision and Control
citation:   Sampathirao, Ajay Kumar and Sopasakis, Pantelis and Bemporad, Alberto and Patrinos, Panagiotis  Distributed solution of stochastic optimal control problems on GPUs.  In: 54th IEEE Conference on Decision and Control, 15-18 December, 2015, Osaka, Japan         (2015)  
document_url: http://eprints.imtlucca.it/2779/1/APG_GPU.pdf