---
_id: '168'
abstract:
- lang: eng
  text: The use of heterogeneous computing resources, such as Graphic Processing Units
    or other specialized coprocessors, has become widespread in recent years because
    of their per- formance and energy efficiency advantages. Approaches for managing
    and scheduling tasks to heterogeneous resources are still subject to research.
    Although queuing systems have recently been extended to support accelerator resources,
    a general solution that manages heterogeneous resources at the operating system-
    level to exploit a global view of the system state is still missing.In this paper
    we present a user space scheduler that enables task scheduling and migration on
    heterogeneous processing resources in Linux. Using run queues for available resources
    we perform scheduling decisions based on the system state and on task characterization
    from earlier measurements. With a pro- gramming pattern that supports the integration
    of checkpoints into applications, we preempt tasks and migrate them between three
    very different compute resources. Considering static and dynamic workload scenarios,
    we show that this approach can gain up to 17% performance, on average 7%, by effectively
    avoiding idle resources. We demonstrate that a work-conserving strategy without
    migration is no suitable alternative.
author:
- first_name: Achim
  full_name: Lösch, Achim
  id: '43646'
  last_name: Lösch
- first_name: Tobias
  full_name: Beisel, Tobias
  last_name: Beisel
- first_name: Tobias
  full_name: Kenter, Tobias
  id: '3145'
  last_name: Kenter
- first_name: Christian
  full_name: Plessl, Christian
  id: '16153'
  last_name: Plessl
  orcid: 0000-0001-5728-9982
- first_name: Marco
  full_name: Platzner, Marco
  id: '398'
  last_name: Platzner
citation:
  ama: 'Lösch A, Beisel T, Kenter T, Plessl C, Platzner M. Performance-centric scheduling
    with task migration for a heterogeneous compute node in the data center. In: <i>Proceedings
    of the 2016 Design, Automation &#38; Test in Europe Conference &#38; Exhibition
    (DATE)</i>. EDA Consortium / IEEE; 2016:912-917.'
  apa: Lösch, A., Beisel, T., Kenter, T., Plessl, C., &#38; Platzner, M. (2016). Performance-centric
    scheduling with task migration for a heterogeneous compute node in the data center.
    <i>Proceedings of the 2016 Design, Automation &#38; Test in Europe Conference
    &#38; Exhibition (DATE)</i>, 912–917.
  bibtex: '@inproceedings{Lösch_Beisel_Kenter_Plessl_Platzner_2016, title={Performance-centric
    scheduling with task migration for a heterogeneous compute node in the data center},
    booktitle={Proceedings of the 2016 Design, Automation &#38; Test in Europe Conference
    &#38; Exhibition (DATE)}, publisher={EDA Consortium / IEEE}, author={Lösch, Achim
    and Beisel, Tobias and Kenter, Tobias and Plessl, Christian and Platzner, Marco},
    year={2016}, pages={912–917} }'
  chicago: Lösch, Achim, Tobias Beisel, Tobias Kenter, Christian Plessl, and Marco
    Platzner. “Performance-Centric Scheduling with Task Migration for a Heterogeneous
    Compute Node in the Data Center.” In <i>Proceedings of the 2016 Design, Automation
    &#38; Test in Europe Conference &#38; Exhibition (DATE)</i>, 912–17. EDA Consortium
    / IEEE, 2016.
  ieee: A. Lösch, T. Beisel, T. Kenter, C. Plessl, and M. Platzner, “Performance-centric
    scheduling with task migration for a heterogeneous compute node in the data center,”
    in <i>Proceedings of the 2016 Design, Automation &#38; Test in Europe Conference
    &#38; Exhibition (DATE)</i>, 2016, pp. 912–917.
  mla: Lösch, Achim, et al. “Performance-Centric Scheduling with Task Migration for
    a Heterogeneous Compute Node in the Data Center.” <i>Proceedings of the 2016 Design,
    Automation &#38; Test in Europe Conference &#38; Exhibition (DATE)</i>, EDA Consortium
    / IEEE, 2016, pp. 912–17.
  short: 'A. Lösch, T. Beisel, T. Kenter, C. Plessl, M. Platzner, in: Proceedings
    of the 2016 Design, Automation &#38; Test in Europe Conference &#38; Exhibition
    (DATE), EDA Consortium / IEEE, 2016, pp. 912–917.'
date_created: 2017-10-17T12:41:24Z
date_updated: 2023-09-26T13:27:00Z
ddc:
- '040'
department:
- _id: '27'
- _id: '518'
- _id: '78'
file:
- access_level: closed
  content_type: application/pdf
  creator: florida
  date_created: 2018-03-21T12:41:55Z
  date_updated: 2018-03-21T12:41:55Z
  file_id: '1541'
  file_name: 168-07459438.pdf
  file_size: 261356
  relation: main_file
  success: 1
file_date_updated: 2018-03-21T12:41:55Z
has_accepted_license: '1'
language:
- iso: eng
page: 912-917
project:
- _id: '1'
  grant_number: '160364472'
  name: SFB 901
- _id: '14'
  grant_number: '160364472'
  name: SFB 901 - Subprojekt C2
- _id: '4'
  name: SFB 901 - Project Area C
- _id: '30'
  grant_number: 01|H11004A
  name: Enabling Heterogeneous Hardware Acceleration using Novel Programming and Scheduling
    Models
publication: Proceedings of the 2016 Design, Automation & Test in Europe Conference
  & Exhibition (DATE)
publisher: EDA Consortium / IEEE
quality_controlled: '1'
status: public
title: Performance-centric scheduling with task migration for a heterogeneous compute
  node in the data center
type: conference
user_id: '15278'
year: '2016'
...
---
_id: '10624'
abstract:
- lang: eng
  text: "The use of heterogeneous computing resources, such as graphics processing
    units or other specialized co-processors, has become widespread in recent years
    because of their performance and energy efficiency advantages. Operating system
    approaches that are limited to optimizing CPU usage are no longer sufficient for
    the efficient utilization of systems that comprise diverse resource types.\r\n\r\nEnabling
    task preemption on these architectures and migration of tasks between different
    resource types at run-time is not only key to improving the performance and energy
    consumption but also to enabling automatic scheduling methods for heterogeneous
    compute nodes.\r\n\r\nThis thesis proposes novel techniques for run-time management
    of heterogeneous resources and enabling tasks to migrate between diverse hardware.
    It provides fundamental work towards future operating systems by discussing implications,
    limitations, and chances of the heterogeneity and introducing solutions for energy-
    and performance-efficient run-time systems. Scheduling methods to utilize heterogeneous
    systems by the use of a centralized scheduler are presented that show benefits
    over existing approaches in varying case studies."
author:
- first_name: Tobias
  full_name: Beisel, Tobias
  last_name: Beisel
citation:
  ama: 'Beisel T. <i>Management and Scheduling of Accelerators for Heterogeneous High-Performance
    Computing</i>. Berlin: Logos Verlag Berlin GmbH; 2015.'
  apa: 'Beisel, T. (2015). <i>Management and Scheduling of Accelerators for Heterogeneous
    High-Performance Computing</i>. Berlin: Logos Verlag Berlin GmbH.'
  bibtex: '@book{Beisel_2015, place={Berlin}, title={Management and Scheduling of
    Accelerators for Heterogeneous High-Performance Computing}, publisher={Logos Verlag
    Berlin GmbH}, author={Beisel, Tobias}, year={2015} }'
  chicago: 'Beisel, Tobias. <i>Management and Scheduling of Accelerators for Heterogeneous
    High-Performance Computing</i>. Berlin: Logos Verlag Berlin GmbH, 2015.'
  ieee: 'T. Beisel, <i>Management and Scheduling of Accelerators for Heterogeneous
    High-Performance Computing</i>. Berlin: Logos Verlag Berlin GmbH, 2015.'
  mla: Beisel, Tobias. <i>Management and Scheduling of Accelerators for Heterogeneous
    High-Performance Computing</i>. Logos Verlag Berlin GmbH, 2015.
  short: T. Beisel, Management and Scheduling of Accelerators for Heterogeneous High-Performance
    Computing, Logos Verlag Berlin GmbH, Berlin, 2015.
date_created: 2019-07-10T09:36:58Z
date_updated: 2022-01-06T06:50:48Z
department:
- _id: '78'
- _id: '27'
- _id: '518'
language:
- iso: eng
page: '183'
place: Berlin
project:
- _id: '30'
  grant_number: 01|H11004
  name: Enabling Heterogeneous Hardware Acceleration using Novel Programming and Scheduling
    Models
publication_identifier:
  isbn:
  - 978-3-8325-4155-2
publisher: Logos Verlag Berlin GmbH
status: public
supervisor:
- first_name: Christian
  full_name: Plessl, Christian
  id: '16153'
  last_name: Plessl
  orcid: 0000-0001-5728-9982
title: Management and Scheduling of Accelerators for Heterogeneous High-Performance
  Computing
type: dissertation
user_id: '3118'
year: '2015'
...
---
_id: '1787'
author:
- first_name: Tim
  full_name: Suess, Tim
  last_name: Suess
- first_name: Andrew
  full_name: Schoenrock, Andrew
  last_name: Schoenrock
- first_name: Sebastian
  full_name: Meisner, Sebastian
  last_name: Meisner
- first_name: Christian
  full_name: Plessl, Christian
  id: '16153'
  last_name: Plessl
  orcid: 0000-0001-5728-9982
citation:
  ama: 'Suess T, Schoenrock A, Meisner S, Plessl C. Parallel Macro Pipelining on the
    Intel SCC Many-Core Computer. In: <i>Proc. Int. Symp. on Parallel and Distributed
    Processing Workshops (IPDPSW)</i>. IEEE Computer Society; 2013:64-73. doi:<a href="https://doi.org/10.1109/IPDPSW.2013.136">10.1109/IPDPSW.2013.136</a>'
  apa: Suess, T., Schoenrock, A., Meisner, S., &#38; Plessl, C. (2013). Parallel Macro
    Pipelining on the Intel SCC Many-Core Computer. <i>Proc. Int. Symp. on Parallel
    and Distributed Processing Workshops (IPDPSW)</i>, 64–73. <a href="https://doi.org/10.1109/IPDPSW.2013.136">https://doi.org/10.1109/IPDPSW.2013.136</a>
  bibtex: '@inproceedings{Suess_Schoenrock_Meisner_Plessl_2013, place={Washington,
    DC, USA}, title={Parallel Macro Pipelining on the Intel SCC Many-Core Computer},
    DOI={<a href="https://doi.org/10.1109/IPDPSW.2013.136">10.1109/IPDPSW.2013.136</a>},
    booktitle={Proc. Int. Symp. on Parallel and Distributed Processing Workshops (IPDPSW)},
    publisher={IEEE Computer Society}, author={Suess, Tim and Schoenrock, Andrew and
    Meisner, Sebastian and Plessl, Christian}, year={2013}, pages={64–73} }'
  chicago: 'Suess, Tim, Andrew Schoenrock, Sebastian Meisner, and Christian Plessl.
    “Parallel Macro Pipelining on the Intel SCC Many-Core Computer.” In <i>Proc. Int.
    Symp. on Parallel and Distributed Processing Workshops (IPDPSW)</i>, 64–73. Washington,
    DC, USA: IEEE Computer Society, 2013. <a href="https://doi.org/10.1109/IPDPSW.2013.136">https://doi.org/10.1109/IPDPSW.2013.136</a>.'
  ieee: 'T. Suess, A. Schoenrock, S. Meisner, and C. Plessl, “Parallel Macro Pipelining
    on the Intel SCC Many-Core Computer,” in <i>Proc. Int. Symp. on Parallel and Distributed
    Processing Workshops (IPDPSW)</i>, 2013, pp. 64–73, doi: <a href="https://doi.org/10.1109/IPDPSW.2013.136">10.1109/IPDPSW.2013.136</a>.'
  mla: Suess, Tim, et al. “Parallel Macro Pipelining on the Intel SCC Many-Core Computer.”
    <i>Proc. Int. Symp. on Parallel and Distributed Processing Workshops (IPDPSW)</i>,
    IEEE Computer Society, 2013, pp. 64–73, doi:<a href="https://doi.org/10.1109/IPDPSW.2013.136">10.1109/IPDPSW.2013.136</a>.
  short: 'T. Suess, A. Schoenrock, S. Meisner, C. Plessl, in: Proc. Int. Symp. on
    Parallel and Distributed Processing Workshops (IPDPSW), IEEE Computer Society,
    Washington, DC, USA, 2013, pp. 64–73.'
date_created: 2018-03-26T14:51:05Z
date_updated: 2023-09-26T13:38:05Z
department:
- _id: '27'
- _id: '518'
- _id: '78'
- _id: '63'
doi: 10.1109/IPDPSW.2013.136
language:
- iso: eng
page: 64-73
place: Washington, DC, USA
project:
- _id: '30'
  grant_number: 01|H11004A
  name: Enabling Heterogeneous Hardware Acceleration using Novel Programming and Scheduling
    Models
publication: Proc. Int. Symp. on Parallel and Distributed Processing Workshops (IPDPSW)
publication_identifier:
  isbn:
  - 978-0-7695-4979-8
publisher: IEEE Computer Society
quality_controlled: '1'
status: public
title: Parallel Macro Pipelining on the Intel SCC Many-Core Computer
type: conference
user_id: '15278'
year: '2013'
...
---
_id: '2180'
author:
- first_name: Tobias
  full_name: Beisel, Tobias
  last_name: Beisel
- first_name: Tobias
  full_name: Wiersema, Tobias
  id: '3118'
  last_name: Wiersema
- first_name: Christian
  full_name: Plessl, Christian
  id: '16153'
  last_name: Plessl
  orcid: 0000-0001-5728-9982
- first_name: André
  full_name: Brinkmann, André
  last_name: Brinkmann
citation:
  ama: 'Beisel T, Wiersema T, Plessl C, Brinkmann A. Programming and Scheduling Model
    for Supporting Heterogeneous Accelerators in Linux. In: <i>Proc. Workshop on Computer
    Architecture and Operating System Co-Design (CAOS)</i>. ; 2012.'
  apa: Beisel, T., Wiersema, T., Plessl, C., &#38; Brinkmann, A. (2012). Programming
    and Scheduling Model for Supporting Heterogeneous Accelerators in Linux. <i>Proc.
    Workshop on Computer Architecture and Operating System Co-Design (CAOS)</i>.
  bibtex: '@inproceedings{Beisel_Wiersema_Plessl_Brinkmann_2012, title={Programming
    and Scheduling Model for Supporting Heterogeneous Accelerators in Linux}, booktitle={Proc.
    Workshop on Computer Architecture and Operating System Co-design (CAOS)}, author={Beisel,
    Tobias and Wiersema, Tobias and Plessl, Christian and Brinkmann, André}, year={2012}
    }'
  chicago: Beisel, Tobias, Tobias Wiersema, Christian Plessl, and André Brinkmann.
    “Programming and Scheduling Model for Supporting Heterogeneous Accelerators in
    Linux.” In <i>Proc. Workshop on Computer Architecture and Operating System Co-Design
    (CAOS)</i>, 2012.
  ieee: T. Beisel, T. Wiersema, C. Plessl, and A. Brinkmann, “Programming and Scheduling
    Model for Supporting Heterogeneous Accelerators in Linux,” 2012.
  mla: Beisel, Tobias, et al. “Programming and Scheduling Model for Supporting Heterogeneous
    Accelerators in Linux.” <i>Proc. Workshop on Computer Architecture and Operating
    System Co-Design (CAOS)</i>, 2012.
  short: 'T. Beisel, T. Wiersema, C. Plessl, A. Brinkmann, in: Proc. Workshop on Computer
    Architecture and Operating System Co-Design (CAOS), 2012.'
date_created: 2018-04-03T09:18:33Z
date_updated: 2023-09-26T13:40:17Z
department:
- _id: '27'
- _id: '518'
- _id: '78'
keyword:
- funding-enhance
language:
- iso: eng
project:
- _id: '30'
  grant_number: 01|H11004A
  name: Enabling Heterogeneous Hardware Acceleration using Novel Programming and Scheduling
    Models
publication: Proc. Workshop on Computer Architecture and Operating System Co-design
  (CAOS)
quality_controlled: '1'
status: public
title: Programming and Scheduling Model for Supporting Heterogeneous Accelerators
  in Linux
type: conference
user_id: '15278'
year: '2012'
...
---
_id: '2194'
author:
- first_name: Björn
  full_name: Meyer, Björn
  last_name: Meyer
- first_name: Christian
  full_name: Plessl, Christian
  id: '16153'
  last_name: Plessl
  orcid: 0000-0001-5728-9982
- first_name: Jens
  full_name: Förstner, Jens
  id: '158'
  last_name: Förstner
  orcid: 0000-0001-7059-9862
citation:
  ama: 'Meyer B, Plessl C, Förstner J. Transformation of scientific algorithms to
    parallel computing code: subdomain support in a MPI-multi-GPU backend. In: <i>Symp.
    on Application Accelerators in High Performance Computing (SAAHPC)</i>. IEEE Computer
    Society; 2011:60-63. doi:<a href="https://doi.org/10.1109/SAAHPC.2011.12">10.1109/SAAHPC.2011.12</a>'
  apa: 'Meyer, B., Plessl, C., &#38; Förstner, J. (2011). Transformation of scientific
    algorithms to parallel computing code: subdomain support in a MPI-multi-GPU backend.
    <i>Symp. on Application Accelerators in High Performance Computing (SAAHPC)</i>,
    60–63. <a href="https://doi.org/10.1109/SAAHPC.2011.12">https://doi.org/10.1109/SAAHPC.2011.12</a>'
  bibtex: '@inproceedings{Meyer_Plessl_Förstner_2011, title={Transformation of scientific
    algorithms to parallel computing code: subdomain support in a MPI-multi-GPU backend},
    DOI={<a href="https://doi.org/10.1109/SAAHPC.2011.12">10.1109/SAAHPC.2011.12</a>},
    booktitle={Symp. on Application Accelerators in High Performance Computing (SAAHPC)},
    publisher={IEEE Computer Society}, author={Meyer, Björn and Plessl, Christian
    and Förstner, Jens}, year={2011}, pages={60–63} }'
  chicago: 'Meyer, Björn, Christian Plessl, and Jens Förstner. “Transformation of
    Scientific Algorithms to Parallel Computing Code: Subdomain Support in a MPI-Multi-GPU
    Backend.” In <i>Symp. on Application Accelerators in High Performance Computing
    (SAAHPC)</i>, 60–63. IEEE Computer Society, 2011. <a href="https://doi.org/10.1109/SAAHPC.2011.12">https://doi.org/10.1109/SAAHPC.2011.12</a>.'
  ieee: 'B. Meyer, C. Plessl, and J. Förstner, “Transformation of scientific algorithms
    to parallel computing code: subdomain support in a MPI-multi-GPU backend,” in
    <i>Symp. on Application Accelerators in High Performance Computing (SAAHPC)</i>,
    2011, pp. 60–63, doi: <a href="https://doi.org/10.1109/SAAHPC.2011.12">10.1109/SAAHPC.2011.12</a>.'
  mla: 'Meyer, Björn, et al. “Transformation of Scientific Algorithms to Parallel
    Computing Code: Subdomain Support in a MPI-Multi-GPU Backend.” <i>Symp. on Application
    Accelerators in High Performance Computing (SAAHPC)</i>, IEEE Computer Society,
    2011, pp. 60–63, doi:<a href="https://doi.org/10.1109/SAAHPC.2011.12">10.1109/SAAHPC.2011.12</a>.'
  short: 'B. Meyer, C. Plessl, J. Förstner, in: Symp. on Application Accelerators
    in High Performance Computing (SAAHPC), IEEE Computer Society, 2011, pp. 60–63.'
date_created: 2018-04-03T14:55:57Z
date_updated: 2023-09-26T13:44:11Z
department:
- _id: '27'
- _id: '518'
- _id: '15'
- _id: '78'
doi: 10.1109/SAAHPC.2011.12
keyword:
- tet_topic_hpc
language:
- iso: eng
page: 60-63
project:
- _id: '30'
  grant_number: 01|H11004A
  name: Enabling Heterogeneous Hardware Acceleration using Novel Programming and Scheduling
    Models
publication: Symp. on Application Accelerators in High Performance Computing (SAAHPC)
publisher: IEEE Computer Society
quality_controlled: '1'
status: public
title: 'Transformation of scientific algorithms to parallel computing code: subdomain
  support in a MPI-multi-GPU backend'
type: conference
user_id: '15278'
year: '2011'
...
---
_id: '2193'
author:
- first_name: Tobias
  full_name: Beisel, Tobias
  last_name: Beisel
- first_name: Tobias
  full_name: Wiersema, Tobias
  id: '3118'
  last_name: Wiersema
- first_name: Christian
  full_name: Plessl, Christian
  id: '16153'
  last_name: Plessl
  orcid: 0000-0001-5728-9982
- first_name: André
  full_name: Brinkmann, André
  last_name: Brinkmann
citation:
  ama: 'Beisel T, Wiersema T, Plessl C, Brinkmann A. Cooperative multitasking for
    heterogeneous accelerators in the Linux Completely Fair Scheduler. In: <i>Proc.
    Int. Conf. on Application-Specific Systems, Architectures, and Processors (ASAP)</i>.
    IEEE Computer Society; 2011:223-226. doi:<a href="https://doi.org/10.1109/ASAP.2011.6043273">10.1109/ASAP.2011.6043273</a>'
  apa: Beisel, T., Wiersema, T., Plessl, C., &#38; Brinkmann, A. (2011). Cooperative
    multitasking for heterogeneous accelerators in the Linux Completely Fair Scheduler.
    <i>Proc. Int. Conf. on Application-Specific Systems, Architectures, and Processors
    (ASAP)</i>, 223–226. <a href="https://doi.org/10.1109/ASAP.2011.6043273">https://doi.org/10.1109/ASAP.2011.6043273</a>
  bibtex: '@inproceedings{Beisel_Wiersema_Plessl_Brinkmann_2011, title={Cooperative
    multitasking for heterogeneous accelerators in the Linux Completely Fair Scheduler},
    DOI={<a href="https://doi.org/10.1109/ASAP.2011.6043273">10.1109/ASAP.2011.6043273</a>},
    booktitle={Proc. Int. Conf. on Application-Specific Systems, Architectures, and
    Processors (ASAP)}, publisher={IEEE Computer Society}, author={Beisel, Tobias
    and Wiersema, Tobias and Plessl, Christian and Brinkmann, André}, year={2011},
    pages={223–226} }'
  chicago: Beisel, Tobias, Tobias Wiersema, Christian Plessl, and André Brinkmann.
    “Cooperative Multitasking for Heterogeneous Accelerators in the Linux Completely
    Fair Scheduler.” In <i>Proc. Int. Conf. on Application-Specific Systems, Architectures,
    and Processors (ASAP)</i>, 223–26. IEEE Computer Society, 2011. <a href="https://doi.org/10.1109/ASAP.2011.6043273">https://doi.org/10.1109/ASAP.2011.6043273</a>.
  ieee: 'T. Beisel, T. Wiersema, C. Plessl, and A. Brinkmann, “Cooperative multitasking
    for heterogeneous accelerators in the Linux Completely Fair Scheduler,” in <i>Proc.
    Int. Conf. on Application-Specific Systems, Architectures, and Processors (ASAP)</i>,
    2011, pp. 223–226, doi: <a href="https://doi.org/10.1109/ASAP.2011.6043273">10.1109/ASAP.2011.6043273</a>.'
  mla: Beisel, Tobias, et al. “Cooperative Multitasking for Heterogeneous Accelerators
    in the Linux Completely Fair Scheduler.” <i>Proc. Int. Conf. on Application-Specific
    Systems, Architectures, and Processors (ASAP)</i>, IEEE Computer Society, 2011,
    pp. 223–26, doi:<a href="https://doi.org/10.1109/ASAP.2011.6043273">10.1109/ASAP.2011.6043273</a>.
  short: 'T. Beisel, T. Wiersema, C. Plessl, A. Brinkmann, in: Proc. Int. Conf. on
    Application-Specific Systems, Architectures, and Processors (ASAP), IEEE Computer
    Society, 2011, pp. 223–226.'
date_created: 2018-04-03T14:37:14Z
date_updated: 2023-09-26T13:43:48Z
department:
- _id: '27'
- _id: '518'
- _id: '78'
doi: 10.1109/ASAP.2011.6043273
language:
- iso: eng
page: 223-226
project:
- _id: '30'
  grant_number: 01|H11004A
  name: Enabling Heterogeneous Hardware Acceleration using Novel Programming and Scheduling
    Models
publication: Proc. Int. Conf. on Application-Specific Systems, Architectures, and
  Processors (ASAP)
publisher: IEEE Computer Society
quality_controlled: '1'
status: public
title: Cooperative multitasking for heterogeneous accelerators in the Linux Completely
  Fair Scheduler
type: conference
user_id: '15278'
year: '2011'
...