Energy-aware actor-based distributed programming

*Energy-aware actor-based distributed programming*

The SEED program (academic track)

1 Definition
════════════

1.1 Domain and scientific/technical context
───────────────────────────────────────────

  The energy consumption of software infrastructures and applications,
  notably in the domains of the Cloud-Edge-IoT (CEI) continuum and
  AI-intensive software systems is a major challenge for today's
  societies. Controlling the energy consumption is notably crucial in
  two different contexts: (i) the management/optimization of the overall
  energy consumption of large-scale software systems among with
  AI-intensive software systems now occupy a place of choice and (ii)
  the integration of massive numbers and large varieties of small,
  battery powered IoT devices in large-scale distributed systems.

  Currently, energy consumption is typically handled by runtime
  monitoring of components of the software systems running on
  distributed systems, e.g., data centers, grid architectures, sensor
  networks and other distributed cyber-physical infrastructures. Based
  on measured energy consumption and (configured) available energy
  quotas, some scheduling component then dynamically enables
  computations based on energy policies. [SDU+23, HCW+21, CKC+21]

1.2 Scientific/technical challenges
───────────────────────────────────

  Current distributed software/hardware systems almost never provide
  guarantees that future computations (that are typically encapsulated
  in some software component) can be executed given the currently
  available energy.  Furthermore, energy-related information is
  currently almost always expressed at the API and systems-level
  [RWS+24]: energy-related properties therefore cannot be expressed at
  the programming level. Such guarantees require a precise link between
  energy-consuming computations, i.e., distributed programs, energy
  sources (available electricity networks, batteries) and available
  energy quotas determined based through dynamic monitoring.

  Such a link has been proposed for non-distributed programs (through
  program language extensions or transformations [CY17, ZLL15, YVS17] as
  well as static analysis [MSP+21, KKK16]) but is lacking for
  distributed systems. Currently, software systems for distributed
  systems are typically structured in terms of separate programs that
  are deployed on different machines, often using containers [SDU+23],
  and cooperate at runtime. More precise energy policies and
  energy-aware schedulers can thus be defined only in terms of
  specifications that are difficult to link to the set of executing
  programs.

1.3 Considered methods, targeted results and impacts
────────────────────────────────────────────────────

  The main goal of this PhD is the *precise definition and efficient
  enforcement of energy contracts over distributed programs*. We are
  targeting a distributed programming language extension that allows
  required and available energy quotas of programs to be partially
  defined by the programmer and partially provided dynamically by the
  distributed environment in which the program is executed. Required
  energy quotas are then checked using static analysis or dynamic tests
  against the energy available from the execution environment. Energy
  quotas are managed and enforced through contexts that energy-aware
  programs use to store energy-related environmental
  information. Contexts are also harnessed in order to actively manage
  energy-consuming computations using placement of computations,
  optimization of communication, rate limiting, scaling etc.

  Distributed computations are represented and energy contexts
  maintained in the envisioned programming method by means of *software
  actors* [MSD18, HSF16], components that encapsulate computations and
  data that can be flexibly deployed and executed in distributed
  systems.

1.4 Interdisciplinarity aspects
───────────────────────────────

  The topic straddles the domains of software engineering, energy
  efficiency and frugal computing that is useful in numerous other
  domains. It includes strong interdisplinary aspects, notably through
  applications in the fields of data science (e.g., energy optimization
  of data analyses) and the Industry of the Future (e.g., the
  optimization of monolithic batch processes that are currently used
  frequently). These interdisciplinary aspects will be explored, in
  particular, together with the non academic partners of this topic.

1.5 References
──────────────

  • [CY17] Anthony Canino and Yu David Liu. 2017. Proactive and adaptive
    energy-aware programming with mixed typechecking. In Proceedings of
    the 38th ACM SIGPLAN Conference on Programming Language Design and
    Implementation (PLDI 2017). Association for Computing Machinery, New
    York, NY, USA, 217–232. <https://doi.org/10.1145/3062341.3062356>

  • [ZLL15] Haitao Steve Zhu, Chaoren Lin, and Yu David Liu. 2015. A
    programming model for sustainable software. In Proceedings of the
    37th International Conference on Software Engineering - Volume 1
    (ICSE '15). IEEE Press, 767–777.

  • [KKK16] C. H. P. Kim, D. Kroening and M. Kwiatkowska, "Static
    Program Analysis for Identifying Energy Bugs in Graphics-Intensive
    Mobile Apps," 2016 IEEE 24th International Symposium on Modeling,
    Analysis and Simulation of Computer and Telecommunication Systems
    (MASCOTS), London, UK, 2016, pp. 115-124, doi:
    10.1109/MASCOTS.2016.28.

  • [MSP+21] Marantos, C., Salapas, K., Papadopoulos, L. et al. A
    Flexible Tool for Estimating Applications Performance and Energy
    Consumption Through Static Analysis. SN COMPUT. SCI. 2, 21
    (2021). <https://doi.org/10.1007/s42979-020-00405-7>

  • [SDU+23] R. C. Sofia, D. Dykeman, P. Urbanetz, A. Galal and
    D. A. Dave, "Dynamic, Context-Aware Cross-Layer Orchestration of
    Containerized Applications," in IEEE Access, vol. 11,
    pp. 93129-93150, 2023, doi: 10.1109/ACCESS.2023.3307026.

  • [HCW+21] Yongsheng Hao, Jie Cao, Qi Wang, Jinglin Du, Energy-aware
    scheduling in edge computing with a clustering method, Future
    Generation Computer Systems, Volume 117, 2021, Pages 259-272, ISSN
    0167-739X, <https://doi.org/10.1016/j.future.2020.11.029>.

  • [CKC+21] Chaurasia, N., Kumar, M., Chaudhry, R. et al. Comprehensive
    survey on energy-aware server consolidation techniques in cloud
    computing. J Supercomput 77, 11682–11737
    (2021). <https://doi.org/10.1007/s11227-021-03760-1>

  • [RWS+24] Saurabhsingh Rajput, Tim Widmayer, Ziyuan Shang, Maria
    Kechagia, Federica Sarro, and Tushar Sharma. 2024. Enhancing
    Energy-Awareness in Deep Learning through Fine-Grained Energy
    Measurement. ACM Trans. Softw. Eng. Methodol. 33, 8, Article 211
    (November 2024), 34 pages. <https://doi.org/10.1145/3680470>

  • [YVS17] T. -J. Yang, Y. -H. Chen and V. Sze, "Designing
    Energy-Efficient Convolutional Neural Networks Using Energy-Aware
    Pruning," 2017 IEEE Conference on Computer Vision and Pattern
    Recognition (CVPR), Honolulu, HI, USA, 2017, pp. 6071-6079, doi:
    10.1109/CVPR.2017.643.

  • [MSD18] Myter, F., Scholliers, C., De Meuter, W. (2018). Parallel
    and Distributed Web Programming with Actors. In: Ricci, A., Haller,
    P. (eds) Programming with Actors. Lecture Notes in Computer
    Science(), vol 10789. Springer,
    Cham. <https://doi.org/10.1007/978-3-030-00302-9_1>

  • [HSF16] Hayduk, Y., Sobe, A., Felber, P. (2016). Enhanced Energy
    Efficiency with the Actor Model on Heterogeneous Architectures. In:
    Jelasity, M., Kalyvianaki, E. (eds) Distributed Applications and
    Interoperable Systems. DAIS 2016. Lecture Notes in Computer
    Science(), vol 9687. Springer,
    Cham. <https://doi.org/10.1007/978-3-319-39577-7_1>

2 Partners and study periods
════════════════════════════

2.1 Supervisors and study periods
─────────────────────────────────

  • *IMT Atlantique*: [_Prof. Mario Südholt_], IMT Atlantique, Nantes,
    France

    The PhD student will stay 2 years at IMT Atlantique.

  • *International partner*: [_Prof. Coen De Roover_] and
    [_Prof. Wolfgang De Meuter_], Vrije Universiteit Brussel, Brussels,
    Belgium

    The PhD student will stay 1 year at VUB.

  • *Industrial partner(s)* for short-term visits have not yet been
    determined. However, cooperations with non-academic partners on
    similar topics will be harnessed: e.g., with Nantes university
    hospital for healthcare applications and Zensor SA in Brussels for
    distributed health monitoring of industrial assets.

[_Prof. Mario Südholt_] <http://sudholt.eu>

[_Prof. Coen De Roover_] <http://soft.vub.ac.be/~cderoove/>

[_Prof. Wolfgang De Meuter_]
<https://researchportal.vub.be/nl/persons/wolfgang-de-meuter>

2.2 Hosting organizations
─────────────────────────

2.2.1 IMT Atlantique
╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌

  [_IMT Atlantique_], internationally recognized for the quality of its
  research, is a leading French technological university under the
  supervision of the Ministry of Industry and Digital Technology.  IMT
  Atlantique maintains privileged relationships with major national and
  international industrial partners, as well as with a dense network of
  SMEs, start-ups, and innovation networks. With 290 permanent staff,
  2,200 students, including 300 doctoral students, IMT Atlantique
  produces 1,000 publications each year and raises 18€ million in
  research funds.

[_IMT Atlantique_] <https://www.imt-atlantique.fr/en>

2.2.2 Vrije Universiteit Brussel
╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌

  The VUB is an 'Urban Engaged University': our university connects. We
  forge links between our people - students, staff and stakeholders -,
  society and the wider world. We have always been urban because of our
  location in Brussels, capital of Belgium and Europe. And we are
  engaged in accordance with our principles of radical humanism and
  open, critical thinking. Through excellent research and qualitative
  education on a human scale, the VUB wants to make an active and
  engaged contribution to a better society.

  Are you as concerned as we are about better living conditions, equal
  rights, peace, freedom of speech, a better environment? Well, you are
  in luck: as VUBer you could not be better placed to have an impact on
  these issues, thanks to the knowledge you acquire in your studies or
  your expertise as a researcher. But also because there are dozens of
  projects at the VUB where you can contribute to a better
  world. Initiated by people like you, people the world needs. Welcome
  to The World Needs You.

2.2 Hosting organizations
─────────────────────────

2.2.1 IMT Atlantique
╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌

  [_IMT Atlantique_], internationally recognized for the quality of its
  research, is a leading French technological university under the
  supervision of the Ministry of Industry and Digital Technology.  IMT
  Atlantique maintains privileged relationships with major national and
  international industrial partners, as well as with a dense network of
  SMEs, start-ups, and innovation networks. With 290 permanent staff,
  2,200 students, including 300 doctoral students, IMT Atlantique
  produces 1,000 publications each year and raises 18€ million in
  research funds.

[_IMT Atlantique_] <https://www.imt-atlantique.fr/en>

2.2.2 Vrije Universiteit Brussel
╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌

  The VUB is an 'Urban Engaged University': our university connects. We
  forge links between our people - students, staff and stakeholders -,
  society and the wider world. We have always been urban because of our
  location in Brussels, capital of Belgium and Europe. And we are
  engaged in accordance with our principles of radical humanism and
  open, critical thinking. Through excellent research and qualitative
  education on a human scale, the VUB wants to make an active and
  engaged contribution to a better society.

  Are you as concerned as we are about better living conditions, equal
  rights, peace, freedom of speech, a better environment? Well, you are
  in luck: as VUBer you could not be better placed to have an impact on
  these issues, thanks to the knowledge you acquire in your studies or
  your expertise as a researcher. But also because there are dozens of
  projects at the VUB where you can contribute to a better
  world. Initiated by people like you, people the world needs. Welcome
  to The World Needs You.

Proficiency in programming languages, distributed systems and English.

Knowledge in Green Computing desirable