Please use this identifier to cite or link to this item:
http://hdl.handle.net/1942/31363
Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | NEMETH, Balazs | - |
dc.contributor.author | HABER, Tom | - |
dc.contributor.author | LIESENBORGS, Jori | - |
dc.contributor.author | LAMOTTE, Wim | - |
dc.date.accessioned | 2020-07-01T11:59:25Z | - |
dc.date.available | 2020-07-01T11:59:25Z | - |
dc.date.issued | 2020 | - |
dc.date.submitted | 2020-06-29T10:48:55Z | - |
dc.identifier.citation | Proceedings 20th IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, p. 752 -759 | - |
dc.identifier.isbn | 9781728160955 | - |
dc.identifier.uri | http://hdl.handle.net/1942/31363 | - |
dc.description.abstract | As scientists are designing increasingly complex and intricate models, the prominent way today to achieve acceptable execution time without sacrificing accuracy is through parallel computing. These techniques can improve execution time either on the level of the optimization methods or on the level of the model evaluations. This paper outlines an automatic par-allelization approach for the latter. Processor specific procedures with embedded communication primitives are generated from static schedules produced by an evolutionary algorithm. These are passed to an optimizing compiler to avoid the overhead of typical task runtime systems. The two key insights are that the parallel structure of probabilistic models is revealed when the data is combined with the model and that static schedules can be combined into more robust schedules that can deal with varying load imbalance. For this, LogP model parameters and execution time of each computational task are measured and fed into a discrete event simulator to estimate the running time on the target parallel system. Performance is evaluated with three pharmacological models with different characteristics. The first model lacks enough exploitable parallelism while up to approximately 6x and 8x improvements are achieved for the other models. Compared to a theoretical system with infinite processors and no communication delay, this equates to exploiting 66% and 99% of the available parallelism. Performance improves even when load imbalance varies. | - |
dc.language.iso | en | - |
dc.publisher | IEEE COMPUTER SOC | - |
dc.relation.ispartofseries | BMS | - |
dc.rights | 2020 IEEE | - |
dc.subject.other | Index Terms-High Performance Computing | - |
dc.subject.other | Descriptive Language | - |
dc.subject.other | Probabilistic Modeling | - |
dc.subject.other | Automatic Parallelization | - |
dc.subject.other | Dataflow | - |
dc.subject.other | LogP model | - |
dc.subject.other | Simulation | - |
dc.subject.other | Evolutionary Algorithms | - |
dc.subject.other | Scheduling | - |
dc.subject.other | Load Imbalance | - |
dc.title | Automatic Parallelization of Probabilistic Models with Varying Load Imbalance | - |
dc.type | Proceedings Paper | - |
local.bibliographicCitation.authors | Lefevre, Laurent | - |
local.bibliographicCitation.authors | Varela, Carlos A. | - |
local.bibliographicCitation.authors | Pallis, George | - |
local.bibliographicCitation.authors | Toosi, Adel N. | - |
local.bibliographicCitation.authors | Rana, Omer | - |
local.bibliographicCitation.authors | Buyya, Rajkumar | - |
local.bibliographicCitation.conferencedate | 2020 11-14 May | - |
local.bibliographicCitation.conferencename | 2020 20th IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing (CCGRID) | - |
local.bibliographicCitation.conferenceplace | Australia melbourne | - |
dc.identifier.epage | 759 | - |
dc.identifier.spage | 752 | - |
local.bibliographicCitation.jcat | C1 | - |
local.publisher.place | 10662 LOS VAQUEROS CIRCLE, PO BOX 3014, LOS ALAMITOS, CA 90720-1264 USA | - |
local.type.refereed | Refereed | - |
local.type.specified | Proceedings Paper | - |
local.relation.ispartofseriesnr | CFP20276-ART | - |
dc.identifier.doi | 10.1109/CCGrid49817.2020.00-14 | - |
dc.identifier.isi | WOS:000649540400079 | - |
local.provider.type | - | |
local.bibliographicCitation.btitle | Proceedings 20th IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing | - |
local.uhasselt.uhpub | yes | - |
local.uhasselt.international | no | - |
item.contributor | NEMETH, Balazs | - |
item.contributor | HABER, Tom | - |
item.contributor | LIESENBORGS, Jori | - |
item.contributor | LAMOTTE, Wim | - |
item.validation | ecoom 2022 | - |
item.fulltext | With Fulltext | - |
item.accessRights | Restricted Access | - |
item.fullcitation | NEMETH, Balazs; HABER, Tom; LIESENBORGS, Jori & LAMOTTE, Wim (2020) Automatic Parallelization of Probabilistic Models with Varying Load Imbalance. In: Proceedings 20th IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing, p. 752 -759. | - |
Appears in Collections: | Research publications |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
609500a752.pdf Restricted Access | Published version | 665.92 kB | Adobe PDF | View/Open Request a copy |
SCOPUSTM
Citations
1
checked on Sep 5, 2020
WEB OF SCIENCETM
Citations
2
checked on Oct 12, 2024
Page view(s)
86
checked on Sep 7, 2022
Download(s)
12
checked on Sep 7, 2022
Google ScholarTM
Check
Altmetric
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.