2025Activity reportProject-Team‌​‌MAGELLAN

7.1.1 Tansiv

• Name:
  Time-Accurate​​​‌ Network Simulation Interconnecting Vms​
• Keywords:
  Operating system, Virtualization,​‌ Cloud, Simulation, Cybersecurity
• Functional​​ Description:
  Tansiv: Time-Accurate Network​​​‌ Simulation Interconnecting Virtual machines​ (VMs). Tansiv is a​‌ novel way to run​​ an unmodified distributed application​​​‌ on top of a​ simulated network in a​‌ time accurate and stealth​​ way. To this aim,​​​‌ the VMs execution is​ coordinated (interrupted and restarted)​‌ in order to garantee​​ accurate arrival and transfer​​​‌ of network packets while​ ensuring realistic time flow​‌ within the VMs. The​​ project can leverage several​​​‌ frameworks for simulating the​ data (SimGrid or ns-3)​‌ and several virtualization solutions​​ to encapsulate the application,​​​‌ intercept the network traffic​ and enforce the interruption​‌ decision (Qemu in emulation​​ mode, KVM and Xen​​​‌ with hardware-accelerated virtualization). Tansiv​ can be used in​‌ various situations: malware analysis​​ (e.g. to defeat malware​​​‌ evasion technique based on​ network timing measures) or​‌ analysis of an application​​ on a geo-distributed context.​​​‌
• Contact:
  Louis Rilling
• Partner:​
  DGA-MI

7.1.2 SimGrid

• Keywords:​‌
  Large-scale Emulators, Grid Computing,​​ Distributed Applications
• Scientific Description:​​​‌

  SimGrid is a toolkit​ that provides core functionalities​‌ for the simulation of​​ distributed applications in heterogeneous​​​‌ distributed environments. The simulation​ engine uses algorithmic and​‌ implementation techniques toward the​​ fast simulation of large​​​‌ systems on a single​ machine. The models are​‌ theoretically grounded and experimentally​​ validated. The results are​​​‌ reproducible, enabling better scientific​ practices.

  Its models of​‌ networks, cpus and disks​​ are adapted to (Data)Grids,​​​‌ P2P, Clouds, Fogs, Clusters​ and HPC, allowing multi-domain​‌ studies. It can be​​ used either to simulate​​​‌ algorithms and prototypes of​ applications, or to emulate​‌ real MPI applications through​​ the virtualization of their​​​‌ communication, or to formally​ assess algorithms and applications​‌ that can run in​​ the framework.

  The formal​​​‌ verification module explores all​ possible message interleavings in​‌ the application, searching for​​ states violating the provided​​​‌ properties. This tool can​ be used to assess​‌ safety properties over arbitrary​​ and legacy codes, thanks​​ to a system-level introspection​​​‌ tool that provides a‌ finely detailed view of‌​‌ the running application to​​ the model checker. This​​​‌ can for example be‌ leveraged to verify arbitrary‌​‌ MPI code written in​​ C/C++/Fortran.

• Functional Description:
  SimGrid​​​‌ is a simulation toolkit‌ that provides core functionalities‌​‌ for the simulation of​​ distributed applications in large​​​‌ scale heterogeneous distributed environments.‌
• Release Contributions:

  Breaking the‌​‌ seal: v4.0 was not​​ the final release.

  *​​​‌ Allow one to unseal‌ netzones to modify the‌​‌ platform even after the​​ simulation start. * The​​​‌ model-checker can now report‌ memory race conditions (see‌​‌ tutorial). * Pip builds​​ should now work out​​​‌ of the box. *‌ (+ the usual bug‌​‌ fixes overall, and improvements​​ to the Java/Python bindings).​​​‌

• News of the Year:‌

  There were 2 major‌​‌ releases in 2025. We​​ released v4.0 in March,​​​‌ embodying 10 years of‌ development. This turns SimGrid‌​‌ into a mature and​​ stable research instrument. The​​​‌ users can easily extend‌ this tool to adapt‌​‌ it to their specific​​ research, while trusting its​​​‌ software implementation. Release v4.1‌ was published in November,‌​‌ showing that the development​​ did not stall even​​​‌ if the framework is‌ mostly in maintenance mode.‌​‌ The performance simulation mode​​ was extended to allow​​​‌ modifications of the platform‌ topology during the simulation.‌​‌

  Beyond these technical development,​​ most of the scientific​​​‌ work occurred in the‌ embedded model checker. We‌​‌ first worked on improving​​ the performance of the​​​‌ verification process, and fixed‌ some bugs. Our preferred‌​‌ algorithm (ODPOR reduction +​​ Best-first exploration to enable​​​‌ random walk) is now‌ much faster and consumes‌​‌ much less memory, being​​ 5x faster on small​​​‌ scenarios. But since performance‌ is now linear with‌​‌ the number of states,​​ while it used to​​​‌ be polynomial, 5x faster‌ is the lowest performance‌​‌ boost you can expect​​ from the new version.​​​‌ The number of states‌ to explore for a‌​‌ given scenario is still​​ the same (ODPOR was​​​‌ not improved), but we‌ now can explore these‌​‌ states much faster.

  In​​ addition, we added the​​​‌ ability to check for‌ race conditions in the‌​‌ user code. This feature​​ relies on a specific​​​‌ pass to the clang‌ LLVM compiler, to instrument‌​‌ the memory accesses. We​​ use this feature for​​​‌ teaching purposes at our‌ institutions.

  We are working‌​‌ on a parallel explorer​​ leveraging all cores to​​​‌ accelerate the exploration, but‌ unfortunately, we did not‌​‌ manage to find all​​ the bugs in our​​​‌ parallel explorer yet. We‌ are working on verifying‌​‌ SimGrid with itself to​​ hammer this bug out.​​​‌

• URL:
  https://­simgrid.­org/
• Publication:
  hal-04909441‌
• Contact:
  Martin Quinson
• Participants:‌​‌
  Mathieu Laurent, Anne-Cécile Orgerie,​​ Arnaud Legrand, Augustin Degomme,​​​‌ Arnaud Giersch, Frédéric Suter,‌ Martin Quinson, Samuel Thibault‌​‌
• Partners:
  CNRS, ENS Rennes​​

7.1.3 EnOSlib

• Keywords:
  Distributed​​​‌ Applications, Distributed systems, Evaluation,‌ Grid Computing, Cloud computing,‌​‌ Experimentation, Reproducibility, Linux, Virtualization​​
• Functional Description:

  EnOSlib is​​​‌ a library to help‌ you with your distributed‌​‌ application experiments on bare-metal​​ testbeds. The main parts​​​‌ of your experiment logic‌ is made reusable by‌​‌ the following EnOSlib building​​​‌ blocks:

  - Reusable infrastructure​ configuration: The provider abstraction​‌ allows you to run​​ your experiment on different​​​‌ environments (locally with Vagrant,​ Grid’5000, Chameleon, IoT-LAB and​‌ more) - Reusable software​​ provisioning: In order to​​​‌ configure your nodes, EnOSlib​ exposes different APIs with​‌ different level of expressivity​​ - Reusable services: Install​​​‌ common services such as​ Docker, monitoring stacks, network​‌ emulation... - Reusable experiment​​ facilities: Tasks help you​​​‌ to iterate faster on​ your experimentation workflow

  EnOSlib​‌ is designed for experimentation​​ purpose: benchmark in a​​​‌ controlled environment, academic validation​ …

• Release Contributions:

  To​‌ reduce dependencies, the default​​ pip package no longer​​​‌ includes Jupyter support.

  Add​ support for Ansible 8,​‌ 9 and 10

• URL:​​
  https://­discovery.­gitlabpages.­inria.­fr/­enoslib/
• Publications:
  hal-01664515,​​​‌ hal-01689726
• Contact:
  Mathieu Simonin​
• Participants:
  Mathieu Simonin, 6​‌ anonymous participants

8.1.1 Fog/cloud​ storage services

Participants: Shadi​‌ Ibrahim, Marc Tranzer​​, Mohammad Rizk,​​​‌ Armel Jeatsa Toulepi.​

Replication has been successfully​‌ deployed and practiced to​​ ensure high data availability​​​‌ in large-scale distributed storage​ systems. However, with the​‌ relentless growth of generated​​ and collected data, replication​​​‌ has become expensive not​ only in terms of​‌ storage cost, but also​​ in terms of network​​​‌ cost, hardware cost and​ energy cost. While already​‌ popular for archived data​​ in peer-to-peer (P2P) systems​​​‌ and cold data, in​ recent years, erasure coding​‌ (EC) has been progressively​​ performed on the critical​​​‌ path of data access​ and has been has​‌ been increasingly used in​​ practical storage systems such​​​‌ as Ceph and HDFS.​ The main reason for​‌ such an adoption is​​ that EC can provide​​​‌ the same fault tolerance​ guarantee as replication, but​‌ at a much lower​​ storage cost.

In this​​​‌ context, to effectively and​ efficiently adopt EC in​‌ distributed storage systems, we​​ study the effect of​​​‌ data chunk distribution on​ the performance of reads​‌ and data-intensive applications. We​​ also present the design​​​‌ and evaluation of an​ EC-aware block placement that​‌ balances the distribution of​​ data chunks across nodes​​​‌ 18. In addition,​ we continued working on​‌ (i) understanding the tradeoffs​​ between data access performance,​​​‌ repair bandwidth, and energy​ overhead in a CephFS​‌ cluster; and (ii) addressing​​ the challenge of data​​​‌ retrieval in geo-distributed erasure-coded​ storage systems. Two more​‌ detailed articles on these​​ topics are in preparation.​​​‌ We also designed and​ implemented an erasure-coded storage​‌ system that efficiently and​​ effectively integrates Reed-Solomon codes​​​‌ on top of the​ InterPlanetary File System (IPFS).​‌ A thorough paper on​​ this topic is under​​​‌ review.

Another issue related​ to storage services is​‌ in-memory caching. We study​​ how to use compression​​​‌ to increase the effective​ memory capacity, enabling the​‌ node to handle a​​ larger load while ensuring​​​‌ a fast response time.​ A detailed article on​‌ this topic is in​​ preparation.

8.1.2 Fog/cloud infrastructure​​​‌ design

Participants: Shadi Ibrahim​.

Containers, renowned for​‌ their lightweight nature and​​ flexibility, have seen growing​​​‌ adoption for deploying edge​ services such as web​‌ applications. However, existing cloud-oriented​​ container deployment frameworks fail​​ to address the unique​​​‌ challenges of edge environments,‌ including geographical distribution, device‌​‌ heterogeneity, and resource constraints.​​ This oversight leads to​​​‌ suboptimal performance for latency-sensitive‌ edge services like HPC/AI-powered‌​‌ autonomous driving and edge​​ gaming, which demand rapid​​​‌ startup and immediate responsiveness.‌ Our investigation demonstrates that‌​‌ current on-demand image solutions​​ require excessive client-registry communication,​​​‌ resulting in prolonged Round-Trip‌ Time (RTT) - a‌​‌ particularly severe limitation in​​ geographically distributed edge platforms.​​​‌ Furthermore, we observe that‌ the user-space file system‌​‌ (e.g., FUSE), typically employed​​ to handle device heterogeneity,​​​‌ introduces substantial overhead to‌ the native I/O stack.‌​‌ More critically, our findings​​ re- veal that on-demand​​​‌ image solutions exacerbate storage‌ pressure on resource-constrained edge‌​‌ devices. To overcome these​​ challenges, we introduce EDDE,​​​‌ an edge-optimized container deploy-‌ ment framework that redesigns‌​‌ the on-demand image pipeline​​ 16.

Container orchestration​​​‌ systems, such as Kubernetes,‌ streamline containerized application deployment.‌​‌ As more and more​​ applications are being deployed​​​‌ in Kubernetes, there is‌ an increasing need for‌​‌ rescheduling - relocating a​​ running pod to different​​​‌ nodes - due to‌ system upgrades, node failures,‌​‌ and load- balancing optimizations.​​ Live migration, which transfers​​​‌ services from source nodes‌ to target nodes with‌​‌ minimal downtime, is the​​ ideal support for rescheduling.​​​‌ However, implementing live migration‌ for pods that run‌​‌ stateful services is challenging,​​ because Kubernetes manages pods​​​‌ as stateless. First, the‌ current pod's network namespace‌​‌ initialization process causes a​​ mismatch in the network​​​‌ state between the migrated‌ pod and internal containers.‌​‌ Second, migrating the memory​​ state results in extended​​​‌ downtime. Third, Kubernetes operations‌ on pods do not‌​‌ consider preserving the state​​ of the pods. Therefore,​​​‌ we propose KubeSPT to‌ achieve live migration of‌​‌ stateful pods in rescheduling​​ scenarios 8. First,​​​‌ we synchronize the network‌ state of pods and‌​‌ internal containers by controlling​​ packet flow and implement​​​‌ fast service redirection. Second,‌ we introduce a Hot‌​‌ Data and Lazy-Restore method​​ for memory restoration to​​​‌ reduce migration downtime.

8.1.3‌ Fog/cloud networking services

Participants:‌​‌ François Lemercier, Guillaume​​ Pierre, Haraesh Jayasethu​​​‌ Ramachandran, Yann Lainé-Odic‌.

In the context‌​‌ of Industrial IoT applications​​ (i.e. critical response time​​​‌ environments such as manufacturing‌ or monitoring of critical‌​‌ equipment), latency is at​​ the center of a​​​‌ tremendous number of studies‌ aiming to optimize the‌​‌ placement of resources in​​ distributed architectures. To ensure​​​‌ that the quality of‌ service is guaranteed, several‌​‌ solutions exist to reconfigure​​ component placement (migration) and​​​‌ can reduce the overall‌ latency by changing components‌​‌ and routes. However, knowing​​ precisely which component is​​​‌ the source of problematic‌ latency remains scarcely addressed.‌​‌ When taking a decision​​ for a reconfiguration or​​​‌ a migration, which can‌ be triggered due to‌​‌ latency issues, it can​​ be beneficial to check​​​‌ whether the source of‌ the latency can be‌​‌ solved before instantiating a​​ migration or a full​​​‌ reconfiguration. Proper measurement protocols‌ exist but generally refer‌​‌ to specific case studies​​ and would not allow​​​‌ straightforward integration into edge‌ systems. In this context,‌​‌ we have started work​​​‌ to characterize and evaluate​ the different sources of​‌ latency in such environments,​​ before proposing optimization solutions,​​​‌ notably routing-based approaches, to​ reduce latency.

“Clusterize” was​‌ an Artistic / Scientific​​ project driven by François​​​‌ Lemercier and Loïg Nguyen​ . It aimed to​‌ create a fluid and​​ immersive soundscape, without a​​​‌ beginning or an end,​ embodying the dynamics of​‌ a geo-distributed cluster. This​​ musical installation served as​​​‌ a living mirror of​ digital events, where each​‌ piece of data within​​ the cluster played a​​​‌ role, translated into sounds​ that evolved with the​‌ data flows as well​​ as the gestures and​​​‌ interactions of the audience.​ A dedicated platform was​‌ developed to highlight the​​ artifacts inherent to the​​​‌ execution of a geo-distributed​ Kubernetes cluster, in order​‌ to feed and control​​ the sonic material of​​​‌ the artist participating in​ the Clusterize project.

8.1.4​‌ Reliable fog platforms in​​ resource-constrained natural environments

Participants:​​​‌ Guillaume Pierre, Ammar​ Kazem, Matthieu Nicolas​‌, Elsa Mühl,​​ Haraesh Jayasethu Ramachandran.​​​‌

Natural environment observatories allow​ a wide range of​‌ scientists such as biologists,​​ botanists and hydrologists, to​​​‌ observe a zone of​ particular interest from the​‌ points of view of​​ their different scientific disciplines.​​​‌ They follow a data-driven​ approach based on a​‌ variety of sensors and/or​​ actuators deployed in the​​​‌ natural environment, coupled with​ different techniques to report​‌ the produced data to​​ a public or private​​​‌ cloud for further analysis.​ The constraints which stem​‌ from the specificities of​​ such observatories however deviate​​​‌ from traditional IoT use-cases​ deployed in urban environments.​‌ Observatories are often created​​ in remote locations, where​​​‌ energy supply, cellular networking​ coverage and human maintenance​‌ are more challenging than​​ usual. To address these​​​‌ challenges we are designing​ new fog computing platforms​‌ that can process observation​​ data in-situ while respecting​​​‌ the specific constraints of​ these environments 43.​‌

In this context we​​ conducted a survey of​​​‌ 35 observatories in France​ and abroad to better​‌ understand their existing systems​​ and practice as well​​​‌ as avenues for improvement​ using fog computing technologies​‌ 4. We also​​ started addressing the difficult​​​‌ question of designing a​ fog computing cluster from​‌ a hardware and software​​ point of view that​​​‌ has sufficient processing capacity​ to handle a pre-defined​‌ set of data processing​​ applications while reducing its​​​‌ energy consumption as much​ as possible. For this​‌ we exploit heterogeneous clusters​​ of single-board computers (Raspberry​​​‌ Pi or equivalent) that​ can offer interesting tradeoffs​‌ between processing power and​​ necessary energy. We presented​​​‌ a first demo of​ our solutions 21,​‌ and a more thorough​​ paper on this topic​​​‌ was submitted.

Photo of​ the installation of LivingFog​‌ in Lete, Nepal

We​ continued the work started​‌ last year to redesign​​ and improve the already-existing​​​‌ LivingFog platform 1.​ The objective is to​‌ specialize it to become​​ a viable platform to​​​‌ host a set of​ data processing applications in​‌ a natural environment observatory.​​ The short-time goals are​​ to help detect sensor​​​‌ failures in order to‌ issue timely alarms. A‌​‌ first version of this​​ platform was deployed in​​​‌ the Kaligandaki observatory in‌ Lete, Nepal (see Figure‌​‌ 2).

Finally, we​​ developed the LivingBench benchmarking​​​‌ tool which is designed‌ to exercise an IoT/Edge‌​‌ computing platform based on​​ a use-case derived from​​​‌ our work on the‌ LivingFog platform 20.‌​‌

8.2.1 Reliable serverless​​​‌ runtimes

Participants: Nikos Parlavantzas‌, Mohamed Cherif Zouaoui‌​‌ Latreche.

Function-as-a-Service (FaaS)​​ is a compelling programming​​​‌ model for developing applications‌ that run on fog‌​‌ infrastructures. FaaS applications are​​ composed of ephemeral, event-triggered​​​‌ functions, which can be‌ flexibly deployed at any‌​‌ location on the Edge-Cloud​​ continuum, can be rapidly​​​‌ triggered, and consume resources‌ only when needed. However,‌​‌ supporting fog applications places​​ stringent requirements on FaaS​​​‌ runtimes. First, the runtimes‌ should support latency objectives‌​‌ for functions, critical for​​ latency-sensitive fog applications. Second,​​​‌ the runtimes should support‌ reducing the energy consumption‌​‌ of the underlying infrastructure,​​ particularly important for resource-constrained​​​‌ fog nodes. A major‌ limitation of current FaaS‌​‌ runtimes is their lack​​ of support for meeting​​​‌ latency and energy consumption‌ requirements.

To address this‌​‌ limitation, we are developing​​ FaaS scheduling approaches based​​​‌ on Reinforcement Learning (RL).‌ Our initial work proposed‌​‌ an RL-based approach for​​ optimizing the placement of​​​‌ FaaS functions within a‌ single cluster. Building on‌​‌ this work, in 2025,​​ we have developed a​​​‌ FaaS control plane that‌ spans across multiple geo-distributed‌​‌ clusters, jointly optimizing latency​​ and energy efficiency under​​​‌ dynamically evolving FaaS workloads.‌ Experiments conducted on Grid'5000‌​‌ allowed us to compare​​ RL-based scheduling methods with​​​‌ baseline scheduling methods and‌ explore the trade-offs between‌​‌ centralised and decentralized RL-based​​ control. An article on​​​‌ this topic is currently‌ under submission.

Another limitation‌​‌ of FaaS runtimes in​​ edge environments with resource-constrained​​​‌ devices is memory waste‌ caused by keeping function‌​‌ sandboxes idle to avoid​​ cold starts. To address​​​‌ this, we applied a‌ mechanism that unloads libraries‌​‌ from idle sandboxes and​​ restores them on demand,​​​‌ achieving significant memory savings‌ and improved warm-start performance‌​‌ 25.

8.2.2 Decentralized​​ resource provisioning for serverless​​​‌ applications

Participants: Nikos Parlavantzas‌, Volodia Parol-Guarino.‌​‌

Application resource provisioning in​​ decentralised environments poses significant​​​‌ challenges. First, resources are‌ typically owned and managed‌​‌ by multiple, independent providers,​​ ranging from individuals and​​​‌ small-scale edge cloud operators‌ to large, traditional cloud‌​‌ providers. All these providers​​ should be incentivized to​​​‌ make their resources available‌ to applications. Second, resources‌​‌ are highly heterogeneous and​​ volatile, which, combined with​​​‌ the dynamism of application‌ workloads, makes it difficult‌​‌ to meet application quality​​ of service (QoS) requirements.​​​‌

We aim to address‌ these challenges in the‌​‌ context of FaaS applications.​​ To this end, we​​​‌ have proposed a market-based‌ approach, called GIRAFF, for‌​‌ placing FaaS applications in​​ multi-provider fog infrastructures, along​​​‌ with an open-source implementation.‌ In our approach, clients‌​‌ submit application placement requests​​ associated with SLAs that​​​‌ specify guarantees over network‌ latency and allocated resources.‌​‌ A marketplace then organizes​​​‌ an auction where fog​ nodes bid on the​‌ SLA to determine the​​ nodes that will host​​​‌ the functions and the​ revenue for fog node​‌ owners. To evaluate our​​ approach, we performed reproducible​​​‌ experiments on the Grid’5000​ testbed at large scale​‌ (up to 663 fog​​ nodes on 60 servers).​​​‌ The evaluation demonstrated GIRAFF's​ effectiveness in optimizing FaaS​‌ application placement 26

In​​ 2025, we extended GIRAFF​​​‌ with a client library​ for building fog-native FaaS​‌ applications. The library enables​​ application owners to manage​​​‌ scaling, control costs, and​ handle resource shortages.

8.2.3​‌ Simulation of FaaS applications​​ at the edge

Participants:​​​‌ Nikos Parlavantzas, François​ Lemercier, Sayah Mahfoud​‌ Abd El Ali.​​

Serverless computing is an​​​‌ increasingly popular model for​ delivering cloud services, isolating​‌ developers from infrastructure management​​ and allowing them to​​​‌ focus on application logic.​ At its core, the​‌ Function-as-a-Service (FaaS) model runs​​ short-lived, event-triggered functions that​​​‌ consume resources only during​ execution.

Applying FaaS to​‌ edge computing has gained​​ attention, as it reduces​​​‌ resource usage on constrained​ edge nodes and supports​‌ IoT applications and virtual​​ network functions. However, few​​​‌ simulation tools exist for​ executing FaaS applications in​‌ edge environments. In this​​ context, we benchmarked existing​​​‌ solutions and modified a​ simulator to include an​‌ energy model and a​​ use case tailored to​​​‌ our research project.

8.2.4​ Dynamic platform adaptation for​‌ Federated Learning

Participants: Cédric​​ Tedeschi, Mathis Valli​​​‌, Minh Thong Le​ Viet.

Federated Learning​‌ (FL) has emerged as​​ a paradigm shift enabling​​​‌ heterogeneous clients and devices​ to collaborate on training​‌ a shared global model​​ while preserving the privacy​​​‌ of their local data.​ However, a common yet​‌ impractical assumption in existing​​ FL approaches is that​​​‌ the deployment environment is​ static, which is rarely​‌ true in heterogeneous and​​ highly-volatile environments like the​​​‌ Edge-Cloud Continuum, where FL​ is typically executed. While​‌ most of the current​​ FL approaches process data​​​‌ in an online fashion,​ and are therefore adaptive​‌ by nature, they only​​ support adaptation at the​​​‌ ML/DL level (e.g., through​ continual learning to tackle​‌ data and concept drift),​​ putting aside the effects​​​‌ of system variance. Moreover,​ the study and validation​‌ of FL approaches strongly​​ rely on simulations, which,​​​‌ although informative, tends to​ overlook the real-world complexities​‌ and dynamics of actual​​ deployments, in particular with​​​‌ respect to changing network​ conditions, varying client resources,​‌ and security threats.

We​​ made a first step​​​‌ to address these challenges.​ We investigated the shortcomings​‌ of traditional, static FL​​ models and identified areas​​​‌ of adaptation to tackle​ real-life deployment challenges. We​‌ devised a set of​​ design principles for FL​​​‌ systems that can smartly​ adjust their strategies for​‌ aggregation, communication, privacy, and​​ security in response to​​​‌ changing system conditions. To​ illustrate the benefits envisioned​‌ by these strategies, we​​ presented the results of​​​‌ a set of initial​ experiments on a 25-node​‌ testbed. The experiments, which​​ vary both the number​​​‌ of participating clients and​ the network conditions, show​‌ how existing FL systems​​ are strongly affected by​​ changes in their operational​​​‌ environment. Based on these‌ insights, we proposed a‌​‌ set of take-aways for​​ the FL community, towards​​​‌ further research into FL‌ systems that are not‌​‌ only accurate and scalable​​ but also able to​​​‌ dynamically adapt to the‌ real-world deployment unpredictability.

We‌​‌ also started to invstigate​​ alternate ways where Edge​​​‌ devices learn locally and‌ opportunistically exchange their model‌​‌ so as to build​​ a more precise model.​​​‌ Finding the right trade-off‌ between accuracy of the‌​‌ models obtained within an​​ area (composed of one​​​‌ or several interrelated nodes)‌ and the network traffic‌​‌ generated will constitute the​​ main driver for the​​​‌ solution.

8.2.5 Reliable data‌ stream processing runtimes

Participants:‌​‌ Guillaume Pierre, Khaled​​ Arsalane, Govind Kovilkkatt​​​‌ Panickerveetil, Shadi Ibrahim‌, Tiago Da Silva‌​‌ Barros.

Although data​​ stream processing platforms such​​​‌ as Apache Flink are‌ widely recognized as an‌​‌ interesting paradigm to process​​ IoT data in fog​​​‌ computing platforms, the existing‌ performance models that capture‌​‌ stream processing in geo-distributed​​ environments are theoretical works​​​‌ only, and have not‌ been validated against empirical‌​‌ measurements.

In this context​​ we explored the limitations​​​‌ of stream processing autoscaling‌ systems when dealing with‌​‌ stateful data processing operators​​ in geo-distributed environments where​​​‌ the processing capacity of‌ the compute nodes are‌​‌ heterogeneous. We identified the​​ challenges and limitations of​​​‌ current approaches and proposed‌ directions for addressing them‌​‌ 12. Khaled Arsalane​​ defended his PhD thesis​​​‌ on this topic on‌ December 15th 2025 36‌​‌.

Another issue related​​ to data stream processing​​​‌ focuses on the message‌ brokers such as Apache‌​‌ Kafka that are used​​ to ingest incoming data​​​‌ and reliably distribute them‌ to a data stream‌​‌ processing engine. It is​​ very difficult to right-size​​​‌ such systems because their‌ performance and resource efficiency‌​‌ varies a lot based​​ on many hardware- and​​​‌ configuration-related aspects. Because of‌ their stateful nature they‌​‌ are also difficult to​​ dynamically rescale after their​​​‌ initial deployment, which leads‌ many users to drastically‌​‌ over-provision them “just in​​ case.” We developed a​​​‌ methodology to assist Kafka‌ users in exploring the‌​‌ space of workable configurations​​ and in selecting the​​​‌ optimal one for a‌ given expected workload. A‌​‌ paper on this topic​​ was recently accepted for​​​‌ publication.

Finally, we are‌ investigating how to optimize‌​‌ the allocation of resources​​ and tasks when deploying​​​‌ data streaming processing applications‌ in the Fog, with‌​‌ a focus on latency,​​ throughput, energy consumption, and​​​‌ maximum sustainable throughput.

8.3.1‌​‌ Energy monitoring of Cloud​​ servers

Participants: Anne-Cécile Orgerie​​​‌, Maxime Agusti,‌ Anudipa Mondal, Hayfa‌​‌ Tayeb.

In an​​ effort to raise awareness​​​‌ on the increasing carbon‌ emissions of Cloud computing,‌​‌ the European Corporate Sustainability​​ Reporting Directive effectively requires​​​‌ providers to supply their‌ customers with an assessment‌​‌ of the carbon impact​​ associated with their use.​​​‌ This represents a challenge‌ for bare metal servers,‌​‌ where the deployment of​​ dedicated power meters is​​​‌ often unfeasible at scale.‌ To address this, we‌​‌ presented PPEM-BM 11,​​​‌ a novel sensor-driven modeling​ approach to estimate the​‌ power consumption of bare​​ metal servers using CPU​​​‌ temperature data acquired via​ IPMI. PPEM-BM enhances and​‌ generalizes the existing POWERHEAT​​ method, which correlates CPU​​​‌ temperature with power. Our​ methodology involves training individual​‌ power models, performing cross-evaluation​​ to determine their portability,​​​‌ and then using a​ Learning to Rank (LTR)​‌ model to select the​​ most appropriate pre-trained model​​​‌ for a target server​ based on its hardware​‌ configuration and CPU temperature​​ statistics. An experiment conducted​​​‌ on 1,076 production servers​ at OVHcloud shows that​‌ PPEM-BM demonstrates a significant​​ improvement compared to models​​​‌ based solely on hardware​ profiles. The approach offers​‌ a practical, scalable, and​​ cost-effective solution for hosting​​​‌ providers to monitor energy​ consumption without widespread sensor​‌ deployment.

As the demand​​ for AI-driven workloads increases,​​​‌ the energy consumption of​ Graphics Processing Units (GPUs)​‌ devices has come under​​ intense scrutiny, particularly in​​​‌ hyperscale data centers where​ large numbers of accelerators​‌ are centralized and leased​​ to diverse clients. In​​​‌ the context of cloud​ hyperscalers, GPUs power monitoring​‌ presents several challenges that​​ vary depending on the​​​‌ product offered. The monitoring​ capabilities of physical devices​‌ may be limited or​​ even absent for some​​​‌ products. However, given the​ substantial energy demands of​‌ GPUs, power monitoring is​​ essential for both cloud​​​‌ providers and clients. Operators​ require tools to manage​‌ power distribution effectively, such​​ as balancing workloads across​​​‌ Power Distribution Units (PDUs),​ while clients need visibility​‌ into power usage to​​ optimize their workloads for​​​‌ energy efficiency. To address​ these challenges, we proposed​‌ methods for estimating the​​ energy consumption of jobs​​​‌ running on GPU devices​ in cloud environments, spanning​‌ from shared and managed​​ offerings like ML-as-a-Service (MLaaS)​​​‌ to less managed products​ (e.g., Infrastructure-as-a-Service (IaaS)) 19​‌. Our models demonstrate​​ the benefits of sharing​​​‌ GPUs for small AI​ workloads, as well as​‌ the current sub-optimal utilization​​ of GPUs in cloud​​​‌ hyperscalers, based on insights​ from an IaaS GPU​‌ cluster.

8.3.2 Estimating carbon​​ footprint of computing infrastructures​​​‌

Participants: Anne-Cécile Orgerie,​ Matthieu Simonin, Matteo​‌ Chancerel.

Attributing the​​ carbon costs of shared​​​‌ ICT infrastructures to its​ end-users is frequently promoted​‌ as a way to​​ encourage awareness of environmental​​​‌ impacts and advocate for​ more sustainable practices. We​‌ xplored the intricacies of​​ this approach by focusing​​​‌ on shared ICT infrastructures​ specifically dedicated to academic​‌ research, several of which​​ having recently introduced carbon​​​‌ intensity values for their​ users 17. This​‌ scenario serves as a​​ practical case study for​​​‌ examining the methodologies and​ challenges associated with evaluating​‌ the carbon intensity of​​ shared ICT infrastructures. We​​​‌ explore the choices with​ their limitations, discuss the​‌ objectives behind their implementation​​ of this type of​​​‌ environmental indicator and offer​ actionable insights. This analysis​‌ aims to contribute to​​ the broader discussion on​​​‌ sustainable computing practices and​ the role of environmental​‌ indicators in driving meaningful​​ change.

To reduce their​​​‌ carbon footprint, some Fog​ operators power their nodes​‌ with renewable energy combined​​ with batteries to prevent​​ intermittency. Yet, often they​​​‌ do not consider the‌ embodied carbon footprint of‌​‌ the power infrastructure. Moreover,​​ having 24h/7d uptime for​​​‌ Fog services comes at‌ a high embodied carbon‌​‌ cost and do not​​ favor the right to​​​‌ disconnect. We modeled a‌ Fog infrastructure, geographically spread,‌​‌ powered by solar panels,​​ wind turbines and batteries​​​‌ for its whole life‌ time 13. Its‌​‌ nodes are turned off​​ during the night and​​​‌ can be intermittent during‌ the day. We provide‌​‌ a theoretical framework to​​ analyze the balance between​​​‌ a satisfying uptime of‌ the service, and a‌​‌ small carbon footprint. Our​​ simulation-based evaluation using real​​​‌ traces shows that reaching‌ 95% of uptime during‌​‌ work hours rather than​​ 99% divides by two​​​‌ the size of the‌ powering infrastructure, thus decreasing‌​‌ the carbon footprint of​​ the Fog and providing​​​‌ a usecase for combining‌ both environmental and social‌​‌ concerns of Fog users.​​

Finally, we studied the​​​‌ uniqueness of large scale‌ real-life fine-grained electrical consumption‌​‌ time-series and show its​​ link to privacy threats​​​‌ 7.

8.3.3 Energy-efficiency‌ in distributed systems

Participants:‌​‌ Anne-Cécile Orgerie, Matthieu​​ Silard.

Diverse computing​​​‌ paradigms are continuously emerging,‌ transforming how a device‌​‌ interacts with data and​​ completes tasks. These advancements​​​‌ offer functional benefits, like‌ allowing even basic devices‌​‌ to run powerful applications​​ efficiently by offloading demanding​​​‌ tasks to computing centers‌ with available processing power‌​‌ and storage. This ease​​ of task delegation has​​​‌ led to a boom‌ in the number of‌​‌ users, more powerful and​​ critical applications, and a​​​‌ constant drive to expand‌ computing infrastructures to keep‌​‌ up with intensifying demands.​​ However, this growing capacity​​​‌ and geographical distribution comes‌ with a rising environmental‌​‌ cost. Some solutions are​​ being developed to mitigate​​​‌ the environmental impact, including‌ infrastructure optimization (hardware and‌​‌ design choices), finding dynamic​​ ways for lower energy​​​‌ use (both in the‌ hardware and software), using‌​‌ eco-friendly energy sources, etc.​​ We proposed a state​​​‌ of the art that‌ delves into the evolution‌​‌ of the computing landscape,​​ the environmental challenges it​​​‌ presents, and the solutions‌ being developed to address‌​‌ them 34.

Due​​ to some overprovisioning policies​​​‌ and variable usage, data‌ centers in production can‌​‌ face low average resource​​ utilization. This can result​​​‌ in a waste of‌ underused servers and energy.‌​‌ In this context, virtual​​ machine (VM) consolidation combined​​​‌ with shutdown policies can‌ be a pertinent approach‌​‌ for improving resource utilization​​ and reducing energy consumption​​​‌ of the entire cloud‌ infrastructure. However, VM consolidation‌​‌ requires expensive migration techniques,​​ which can potentially affect​​​‌ performance. Consolidation of workload‌ has been proposed and‌​‌ studied as a core​​ capability since the invention​​​‌ of the Cloud. But‌ after two decades of‌​‌ deployment of Cloud infrastructures,​​ VM consolidation is still​​​‌ rarely used in production‌ for small and large-scale‌​‌ environments. We explored and​​ revisit the potential of​​​‌ savings that can be‌ achieved through a versatile‌​‌ and efficient Virtual Machine​​ consolidation in small and​​​‌ large-scale production infrastructures through‌ usage analysis of two‌​‌ Cloud providers infrastructures 41​​​‌, 5. We​ show that potential benefits​‌ in terms of saved​​ cloud resources and energy​​​‌ usage reduction can occur​ for systems in production.​‌

Energy consumption in 5G/6G​​ wireless networks presents a​​​‌ fundamental challenge, encompassing various​ aspects related to energy,​‌ availability, cost, and environmental​​ concerns. According to the​​​‌ literature, the use of​ sleep modes (SMs) has​‌ emerged as an effective​​ strategy to improve the​​​‌ energy efficiency of base​ stations (BS) within recently​‌ proposed 5G frameworks, while​​ also considering future 6G​​​‌ developments. We proposed a​ comprehensive review to elucidate​‌ what SMs are, the​​ decision process to switch​​​‌ to and from SM​ BSs, and the main​‌ challenges associated with them​​ 6. Additionally, power​​​‌ consumption models are analyzed​ to understand how to​‌ evaluate the performance of​​ SMs. Following an overview​​​‌ of the primary SM​ strategies developed in recent​‌ years, we introduce a​​ matrix model classification of​​​‌ these strategies. This classification​ is based on the​‌ policies and tools implemented​​ for SMs in wireless​​​‌ networks. Finally, based on​ this analysis, we highlight​‌ the most promising strategies​​ and present ideas for​​​‌ future research aimed at​ achieving an energy-efficient and​‌ sustainable use of SMs​​ in 5G/6G wireless networks.​​​‌

The massive integration of​ non-dispatchable Distributed Energy Resource​‌ (DER) and the growing​​ electrification of uses, notably​​​‌ via Electric Vehicles (EVs),​ poses major challenges for​‌ Distribution System Operator (DSO).​​ These evolutions are causing​​​‌ congestion on Medium Voltage​ (MV) networks, and increase​‌ uncertainty of production periods.​​ To address these challenges​​​‌ without costly infrastructure investment,​ we proposed an innovative​‌ demand management solution based​​ on direct, intermediary-free communication​​​‌ between the DSO and​ users 28. Leveraging​‌ existing Advanced Metering Infrastructures​​ (AMIs), this approach broadcasts​​​‌ dynamic incentive signals to​ Electric Vehicle Supply Equipments​‌ (EVSEs), which enables optimized​​ management of energy flows,​​​‌ and thus provides a​ rapid response to network​‌ needs. Our solution has​​ the advantage of using​​​‌ existing power and network​ infrastructures and does not​‌ require to gather private​​ information. In addition, it​​​‌ allows users to choose​ whether or not to​‌ provide their devices as​​ flexible loads. The results​​​‌ show significant potential for​ improving grid stability and​‌ reducing congestion, offering a​​ cost-effective, scalable alternative for​​​‌ smart grids.

Smart grids​ are transforming electricity network​‌ management through the use​​ of Advanced Metering Infrastructure​​​‌ (AMI). They facilitate the​ balance between production and​‌ consumption of electricity by​​ providing real-time data and​​​‌ consumption control for the​ Distribution System Operator (DSO).​‌ However, the implementation details,​​ which are often proprietary,​​​‌ are not readily available.​ We proposed a demo​‌ that presents an innovative​​ visualization platform to simulate​​​‌ a Low Voltage (LV)​ residential network of 48​‌ smart meters 29.​​ The platform provides a​​​‌ visual understanding of the​ communication dynamics in an​‌ AMI by reproducing message​​ exchanges and simulating link​​​‌ disruptions. This demo presents​ the state-of-the-art G3-PLC protocol​‌ stack currently employed in​​ the French AMI deployment​​​‌ and demonstrates the transmission​ of routing and application​‌ messages generated by LOADng​​ and DLMS/COSEM, respectively.

8.3.4​​ Setting limits to reduce​​​‌ the energy consumption of‌ digital infrastructures

Participants: Anne-Cécile‌​‌ Orgerie, Pablo Leboulanger​​, Robin Richard.​​​‌

Ten years after the‌ drafting of the Paris‌​‌ Agreement, the objective to​​ reduce greenhouse gas emissions​​​‌ by half by 2030‌ seems difficult to reach.‌​‌ The impact of ICT​​ technologies is growing year​​​‌ after year. Data centers‌ have a large impact‌​‌ in this domain. As​​ a lot of research​​​‌ is focused on improving‌ energy efficiency of data‌​‌ centers, one could ask​​ if this improvement would​​​‌ suffice to reduce greenhouse‌ gas emissions 45.‌​‌ To study the impact​​ of the energy efficiency​​​‌ on the long term,‌ we proposed to model‌​‌ a data center that​​ has a limited power​​​‌ capacity and is regularly‌ refreshed with new servers,‌​‌ more energy-efficient than the​​ previous ones, to cope​​​‌ with its workload 23‌. Our results explore‌​‌ various growth rates for​​ the load, and the​​​‌ energy efficiency and show‌ that without reduction of‌​‌ usage of data center​​ resources, the improvements in​​​‌ energy efficiency will not‌ be not enough to‌​‌ reach the Paris Agreement's​​ objective.

To reduce the​​​‌ ecological impact of a‌ technology, scientists often focus‌​‌ on energy efficiency issues,​​ ignoring the complex rebound​​​‌ effect generated. We focused‌ on the video transmission‌​‌ technology, and discuss the​​ urgent need to be​​​‌ able to set limits‌ 30. We show‌​‌ that these limits can​​ provoke opposition or circumvention,​​​‌ illustrating the difficulty of‌ the task. Finally, we‌​‌ conclude that the question​​ of limits must be​​​‌ considered as a research‌ problem in its own‌​‌ right, and that it​​ is intrinsically multidisciplinary.

8.3.5​​​‌ Frugal mobile computing

Participants:‌ Martin Quinson, Victorien‌​‌ Elvinger.

Mobile computing​​ is very representative of​​​‌ the ever increasing computational‌ offering. In the recent‌​‌ decades, the hardware capacity​​ has increased exponentially in​​​‌ both computational power, memory‌ and storage space, and‌​‌ display size. This increase​​ is in pace with​​​‌ the software offering that‌ went from scarce text‌​‌ messages to augmented reality​​ and advanced features leveraging​​​‌ embedded neural networks. The‌ only decreasing metric is‌​‌ maybe the battery which​​ was reduced from one​​​‌ week or more to‌ one day or less,‌​‌ despite the steady increase​​ of the battery capacity.​​​‌

In this context, the‌ SmolPhone project is research‌​‌ action exploring other potential​​ designs and their respective​​​‌ advantages for mobile computing.‌ Its practical aspects consist‌​‌ in designing a sort​​ of low-tech smartphone offering​​​‌ some services of a‌ classical smartphone with a‌​‌ one-week battery lifetime. The​​ goal is not to​​​‌ optimize a typical smartphone‌ but rather to reconsider‌​‌ the design space of​​ mobile computing and beyond.​​​‌ We envision the designed‌ solution not as a‌​‌ marketable product but as​​ an effective workbench for​​​‌ future research in the‌ domain of frugal computing.‌​‌

The AEx SmolPhone entered​​ its second year in​​​‌ 2025. On the hardware‌ side, Pierre Nozet (hosted‌​‌ in the TARAN team)​​ designed the first prototypes​​​‌ of the PCB for‌ a future devboard involving‌​‌ heterogeneous multiprocessors and low-power​​​‌ technologies for the screen​ and network. More work​‌ is needed to design​​ a fully functional prototype.​​​‌ On the software side,​ Victorien Elvinger implemented an​‌ adapted runtime that can​​ be used to implement​​​‌ graphical applications targetting limited​ hardware, but still making​‌ it possible to write​​ the applicative logic in​​​‌ a scripting language compiled​ on the board.

8.4.1 Simulating distributed IT​​​‌ systems

Participants: Martin Quinson​.

Our team plays​‌ a major role in​​ the advance of the​​​‌ SimGrid simulator of IT​ systems. This framework has​‌ a major impact on​​ the community. Cited by​​​‌ over 600 scientific papers​ (including 60 PhD thesis​‌ and 150 journal articles),​​ it was used as​​​‌ a scientific instrument by​ thousands of researchers over​‌ the years. This year,​​ we coordinated the release​​​‌ of SimGrid v4.0 46​, concluding ten years​‌ of development. We also​​ published a journal article​​​‌ describing resulting framework and​ its contributions to the​‌ field of distributed system​​ simulation 3.

This​​​‌ year, we used this​ tool to propose an​‌ approach to the problem​​ of measuring the performance​​​‌ of one-sided and collective​ operations on a set​‌ of distributed resources. This​​ problem is not trivial​​​‌ because of the complexity​ of distributed time measurement​‌ on non-synchronized hardware. Simulation​​ is attractive because distributed​​​‌ applications can be executed​ on a controlled set​‌ of resources and in​​ a reproducible manner. Writing​​​‌ a fully calibrated simulator​ for a new software​‌ library represents however a​​ time-and labor-intensive effort that​​​‌ is usually only undertaken​ for well established software​‌ solutions such as the​​ MPI standard. Our approach,​​​‌ detailed in 15 makes​ it possible to study​‌ an existing standard for​​ high-performance computing parallel applications,​​​‌ featuring one-sided communications and​ atomic and collective operations.​‌ We present the several​​ mechanisms that are mandatory​​​‌ to integrate the behavior​ of the real distributed​‌ applications into the simulated​​ platform. Leveraging this simulation​​​‌ framework greatly simplifies this​ endeavor, opening the way​‌ to the study of​​ more distributed middleware and​​​‌ libraries through realistic simulations.​

8.4.2 Understanding software infrastructures​‌

Participants: Léo Cosseron,​​ Martin Quinson, Matthieu​​​‌ Simonin.

One of​ the most challenging infrastructure​‌ to study is the​​ one deployed by malware​​​‌ operators. The difficulties classically​ posed by distributed infrastructures​‌ add to the fact​​ that the infrastructure conceptors​​​‌ explicitly try to evade​ the analysis and/or attack​‌ the analysis system. As​​ an answer, Virtual Machine​​​‌ Introspection (VMI) is used​ by sandbox-based dynamic malware​‌ detection and analysis frameworks​​ to observe malware samples​​​‌ while staying isolated and​ stealthy. Sandbox detection and​‌ evasion techniques based on​​ hypervisor introspection are becoming​​​‌ less of an issue​ since running server and​‌ workstation environments on hypervisors​​ is becoming standard and​​​‌ high-end sandboxes manipulate virtual​ clocks to mask VM​‌ execution pauses caused by​​ VMI. However masking VM​​​‌ execution pauses on multicore​ VMs assumes that pauses​‌ involve all virtual cores,​​ which breaks introspection practices​​​‌ observed with popular tools​ like libVMI. Indeed some​‌ introspection procedures require to​​ pause a virtual core​​ while waiting for an​​​‌ event on another virtual‌ core. Moreover, despite virtualization‌​‌ not being a reliable​​ hint for a malware​​​‌ to be under analysis,‌ single-core VMs are less‌​‌ and less common in​​ production workloads and having​​​‌ only one core is‌ a reliable hint of‌​‌ being analyzed. To lower​​ the VMI detection ability​​​‌ of malware on multi-core‌ VMs, we first proposed‌​‌ several metrics to characterize​​ the side effects caused​​​‌ by a VMI monitor‌ in multi-core VMs. Then‌​‌ we introduced a new​​ strategy to conceal VMI​​​‌ pauses in multi-core guests,‌ and asserted its performance‌​‌ compared to the regular​​ approaches used by sandboxes.​​​‌ Our results show that‌ this new strategy improves‌​‌ the stealthiness of VMI​​ pauses on a multi-core​​​‌ system for some metrics.‌ This work is published‌​‌ in 14.

This​​ work on hiding VMI​​​‌ pauses was also presented‌ in details in the‌​‌ more general framework of​​ Léo Cosseron's PhD thesis​​​‌ defended in December 2025‌ 2, where controlling‌​‌ the flow of time​​ in VMs allows to​​​‌ interconnect VMs with a‌ network simulator. This combination‌​‌ of hardware virtualization and​​ network simulation provides both​​​‌ a tool to analyze‌ real distributed applications in‌​‌ reproducible setups and the​​ infrastructure for a performance-wide​​​‌ stealth network environment in‌ a malware analysis sandboxes.‌​‌

8.4.3 Verifying distributed system's​​ correctness

Participants: Mathieu Laurent​​​‌, Martin Quinson.‌

Assessing the correctness of‌​‌ distributed and parallel applications​​ is notoriously difficult due​​​‌ to the complexity of‌ the concurrent behaviors and‌​‌ the difficulty to reproduce​​ bugs. Model checking and​​​‌ systematic debugging is an‌ appealing solution, but it‌​‌ remains limited to small​​ programs because of the​​​‌ states space explosion problem.‌ In this context, Dynamic‌​‌ Partial Order Reduction (DPOR)​​ techniques have proved successful​​​‌ in exploiting concurrency to‌ verify applications without exploring‌​‌ all their behaviors. However,​​ they may lack of​​​‌ efficiency when tracking non-systematic‌ bugs of real size‌​‌ applications. In this work​​ 22, we suggest​​​‌ two adaptations of the‌ Optimal Dynamic Partial Order‌​‌ Reduction (ODPOR) algorithm with​​ a particular focus on​​​‌ bug finding and explanation.‌ The first adaptation is‌​‌ an out-of-order version called​​ RFS ODPOR which avoids​​​‌ being stuck in uninteresting‌ large parts of the‌​‌ state space. Once a​​ bug is found, the​​​‌ second adaptation takes advantage‌ of ODPOR principles to‌​‌ efficiently find the origins​​ of the bug.

8.4.4​​​‌ Designing problem-specific computing infrastructures‌

Participants: Ophelie Renaud,‌​‌ Martin Quinson.

The​​ Square Kilometre Array Observatory​​​‌ (SKAO) project aims to‌ build the two largest‌​‌ telescope arrays in the​​ world, located in South​​​‌ Africa and Australia. These‌ instruments will generate unprecedented‌​‌ volumes of data, requiring​​ its computing infrastructure (called​​​‌ Science Data Processor –‌ SDP) to operate at‌​‌ terabyte rates under strict​​ energy and performance constraints.​​​‌ Anticipating the computational cost‌ of imaging algorithms is‌​‌ therefore critical for both​​ astronomers and system designers.​​​‌ Yet, testing new algorithmic‌ strategies directly on large‌​‌ HPC platforms is often​​ impractical, due to cost,​​​‌ limited availability, and long‌ development cycles. To address‌​‌ this, we proposed SimSDP​​​‌ 44, 27,​ a rapid prototyping tool​‌ designed to support the​​ development and optimization of​​​‌ radio-astronomy imaging pipelines for​ the future SKAO SDP.​‌ The tool enables early​​ exploration of algorithmic choices​​​‌ and their impact on​ computational performance and energy​‌ consumption without requiring access​​ to large-scale systems.

This​​​‌ year, we illustrated the​ use of SimSDP through​‌ three case studies: (1)​​ modeling different strategies for​​​‌ spectral parallelism (joint vs.​ distributed deconvolution), (2) simulating​‌ pipeline execution on large-scale​​ multi-core, multi-node systems with​​​‌ realistic observation sizes, and​ (3) prototyping pipelines with​‌ real-world data from the​​ NenuFAR instrument. These experiments​​​‌ highlight how SimSDP can​ help astronomers and instrument​‌ designers anticipate the computational​​ cost of imaging algorithms,​​​‌ compare different processing strategies,​ and better align algorithmic​‌ development with the capabilities​​ of future HPC infrastructures​​​‌ 33, 32.​

8.4.5 Supporting experiments on​‌ distributed systems' energy consumption​​

Participants: Matthieu Simonin.​​​‌

EnOSlib 49 software has​ been recently extended to​‌ ease the collection of​​ energy metrics. This allows​​​‌ experimenters to gather energy​ metrics from different sources​‌ being software probes or​​ environnemental metrics offered by​​​‌ SLICES-FR platform. This new​ feature is crucial for​‌ supporting experiment driven research​​ in the field of​​​‌ distributed system experiments on​ energy consumption. This has​‌ been exemplified in various​​ workshops (e.g., 10).​​​‌

8.4.6 Predictive models and​ simulation frameworks for Wi-Fi​‌ networks

Participants: François Lemercier​​, Anne-Cécile Orgerie,​​​‌ Aya Shehata.

We​ proposed a study of​‌ the current consumption of​​ various Medium Access Control​​​‌ (MAC) states for Wi-Fi​ devices, focusing on the​‌ latest IEEE 802.11ax (Wi-​​ Fi 6) and 802.11ac​​​‌ standards 24. Through​ detailed experimental measurements, we​‌ provide new, precise values​​ for the current drawn​​​‌ in MAC states such​ as idle, transmission, reception,​‌ and sleep. We capture​​ real-time current consumption in​​​‌ different scenarios with various​ traffic loads, offering an​‌ understanding of the power​​ usage across different operational​​​‌ modes. The results reveal​ significant differences in power​‌ consumption patterns compared to​​ earlier Wi-Fi standards (e.g.,​​​‌ 802.11n), with notable improvements​ in energy efficiency in​‌ sleep and idle states,​​ but increased current draw​​​‌ in transmission and reception​ states. These findings provide​‌ valuable insights into the​​ energy behavior of recent​​​‌ Wi-Fi devices, offering a​ solid basis for the​‌ simulation of power management​​ optimization strategies in next-generation​​​‌ wireless systems.

8.4.7 Sustainable​ research ecosystem

Participants: Anne-Cécile​‌ Orgerie, Matthieu Simonin​​.

Research on the​​​‌ energy transition raises a​ series of major challenges:​‌ it is a problem-oriented​​ field of research (that​​​‌ of decarbonising the energy​ mix), which must be​‌ resolutely ethical and include​​ principles of social and​​​‌ environmental justice from the​ very beginning... and this​‌ despite the problematic nature​​ of the notion of​​​‌ transition which gives a​ misleading notion of the​‌ process that will lead​​ to desired changes -​​​‌ if there are any.​ Taking this observation as​‌ its starting point, we​​ set out a number​​​‌ of avenues for research​ into key issues 9​‌. It begins by​​ examining the way in​​ which the energy transition​​​‌ is transforming the interplay‌ between actors in the‌​‌ energy sector and the​​ scales at which decisions​​​‌ are taken and implemented.‌ While inviting to take‌​‌ account of the wide​​ variety of national contexts​​​‌ and levels of development,‌ we acknowledge the certainties‌​‌ and uncertainties surrounding the​​ composition of the energy​​​‌ mix. It also stresses‌ the importance of taking‌​‌ account of the externalities​​ of technological choices, particularly​​​‌ in environmental terms. This‌ enables us to draw‌​‌ up a research agenda​​ focusing not only on​​​‌ the spatial aspects of‌ transitions, but also on‌​‌ the ways in which​​ economic models need to​​​‌ incorporate new parameters and‌ new constraints. We also‌​‌ invite to consider how​​ the relationship between energy​​​‌ production and consumption is‌ changing, and how consumption‌​‌ needs to be reduced.​​ In so doing, we​​​‌ defend the idea that‌ research, which is necessarily‌​‌ interdisciplinary and reflexive, has​​ a strong role to​​​‌ play in basing transition‌ action on scientifically sound‌​‌ findings.

We produced a​​ Manifesto from the Perspectives​​​‌ Workshop 25122 entitled “Climate‌ Change: What is Computing's‌​‌ Responsibility?” held March 16-19,​​ 2025 at Schloss Dagstuhl,​​​‌ Germany 40. The‌ workshop provided a forum‌​‌ for world-leading computer scientists​​ and expert consultants on​​​‌ environmental policy and sustainable‌ transition to engage in‌​‌ a critical and urgent​​ conversation about computing's responsibilities​​​‌ in addressing climate change‌ - or more aptly,‌​‌ climate crisis.

We also​​ produced a quiz on​​​‌ the environmental impacts of‌ ICT that was used‌​‌ during several conferences and​​ thematic days 37.​​​‌

We worked with the‌ Paris Observatory to establish‌​‌ their carbon footprint due​​ to ICT infrastructures and​​​‌ devices 38.

We‌ are involved in the‌​‌ Labos1point5 community in which​​ we actively contribute to​​​‌ the implementation of tools‌ and also the animation‌​‌ of the network of​​ laboratories in transition (started​​​‌ in early 2024). The‌ latter offers several opportunities‌​‌ of studies made in​​ conjunction with sociologists.

We​​​‌ have helped several conferences‌ to assess their carbon‌​‌ footprint(IEEEVR - 1500 attendees,​​ SFRBM - 200 attendees,​​​‌ GreenDays and IEEE IC2E‌ - 70 attendee). Initial‌​‌ results show that attending​​ abroad conferences is very​​​‌ carbon intensive: emissions per‌ day and per capita‌​‌ might be 2 orders​​ of magnitude more than​​​‌ the objective set by‌ the Paris agreements. This‌​‌ advocates for finding ways​​ of disseminating scientific results​​​‌ in a sustainable manner‌ in the future.

Défi‌ Inria OVHCloud - FrugalCloud‌​‌

Participants: Anne-Cécile Orgerie,​​ Shadi Ibrahim, Govind​​​‌ Kovilkkatt Panickerveetil, Guillaume‌ Pierre, Maxime Agusti‌​‌, Marc Tranzer.​​

• Title:
  FrugalCloud: Eco-conception​​​‌ de bout en bout‌ d'un cloud pour en‌​‌ réduire les impacts environnementaux​​
• Partner Institution(s):
   

  • Inria, France​​​‌
  • OVHCloud, France
• Date/Duration:
  2021-2025‌
• Additionnal info/keywords:
  The goal‌​‌ of this collaborative framework​​ between the OVHcloud and​​​‌ Inria is to explore‌ new solutions for the‌​‌ design of cloud computing​​ services that are more​​​‌ energy-efficient and environment friendly.‌ Five Inria project-teams are‌​‌ involved in this challenge​​​‌ including Avalon, Inocs, Magellan,​ Spirals, Stack. Members of​‌ the Magellan team contribute​​ to four sub-challenges including​​​‌ (1) Software ecodesign of​ a data stream processing​‌ service; (2) energy-efficient data​​ management; (3) observation of​​​‌ bare metal co-location platforms​ and proposal of energy​‌ reduction catalogues and models;​​ and (4) modelling and​​​‌ designing a framework and​ its environmental Gantt Chart​‌ to manage physical and​​ logical levers.

Défi Inria​​​‌ Hive - Alvearium

Participants:​ Shadi Ibrahim, Mohammad​‌ Rizk, Quentin Acher​​, Armel Jeatsa Toulepi​​​‌.

• Title:
  Alvearium:​ Large Scale Secure and​‌ Reliable Peer-to-Peer Cloud Storage.​​
• Partner Institution(s):
   

  • Inria, France​​​‌
  • Hive, France
• Date/Duration:
  2023-2026​
• Additionnal info/keywords:
  The goal​‌ of this collaborative framework​​ between Hive and Inria​​​‌ is to explore new​ solutions for the design​‌ and realization of large​​ scale secure and reliable​​​‌ Peer-to-Peer Cloud storage. Four​ Inria project-teams are involved​‌ in this challenge including​​ COAST, Magellan, WIDE, COATI.​​​‌ Members of the Myriads​ team will contribute to​‌ two axes. Specifically, the​​ Magellan team coordinates the​​​‌ axis on reliable and​ cost-efficient data placement and​‌ repair in P2P storage​​ over immutable data; and​​​‌ contributes to the axis​ on the management of​‌ mutable data over P2P​​ storage.

Défi Inria Hive​​​‌ - Cupseli

Participants: Shadi​ Ibrahim.

• Title:
  Cupseli​‌: Collaborative Unified Platform​​ for a Scalable and​​​‌ Efficient Learning Infrastructure.
• Partner​ Institution(s):
   

  • Inria, France
  • Hive,​‌ France
• Date/Duration:
  2025-2029
• Additionnal​​ info/keywords:
  The goal of​​​‌ this collaborative framework between​ Hive and Inria is​‌ to to demonstrate that​​ even the most demanding​​​‌ AI and Big Data​ applications can run efficiently​‌ on heterogeneous, distributed, and​​ volatile resources — while​​​‌ maintaining accuracy, ensuring privacy,​ and reducing environmental impact.​‌ Eleven Inria project-teams are​​ involved in this challenge​​​‌ including ARGO, MIMOVE, COAST,​ MAGELLAN, STACK, WIDE, OCKHAM,​‌ COATI, NEO, TADAAM and​​ TOPAL. Members of the​​​‌ MAGELLAN team will contribute​ to the axis on​‌ large-scale computing on intermittent​​ resources.

10.1.1 Associate Teams in​‌ the framework of an​​ Inria International Lab or​​​‌ in the framework of​ an Inria International Program​‌

10.2.1‌​‌ Digital Europe

Participants: Cédric​​ Tedeschi, Guillaume Pierre​​​‌.

• Title: ACHIEVE: Advanced‌ Cloud and High-performance computing‌​‌ Education for a Valiant​​ Europe
• Partner Institution(s):
  • EIT​​​‌ Digital
  • EIT Digital Spain‌
  • Universita Degli Studi di‌​‌ Trento
  • Middle East Technical​​ University
  • Université de Rennes​​​‌
  • Kungliga Tekniska Hoegskolan
  • Universitate‌ Babes Bolyai
  • Aalto University‌​‌
  • Politecnico di Milano
  • University​​ of Novi Sad
  • Evolutionary​​​‌ Archetypes Consulting
  • Techvalley Management‌
  • Infineon
  • Odtu Teknokent
  • Scientific‌​‌ and Technological Research center​​ council of Turkiye
  • Research​​​‌ Institute of Sweden
  • Amazon‌ Web Services
• Date/Duration: 2024-2028‌​‌
• Additionnal info/keywords: ACHIEVE aims​​ to implement a double-degree​​​‌ master's program focusing on‌ Cloud and Networking Infrastructure‌​‌ with specializations that align​​ with the strategic development​​​‌ of advanced, green, and‌ efficient HPC systems. Additionally,‌​‌ it includes a minor​​ in Innovation and Entrepreneurship.​​​‌ This program will be‌ collaboratively designed and delivered‌​‌ by eight higher education​​ institutions from six different​​​‌ countries, in partnership with‌ major industry associations and‌​‌ companies specializing in cloud​​ computing and HPC, along​​​‌ with an innovative SME‌ for educational program delivery,‌​‌ a non-profit association, and​​ EIT Digital, a leader​​​‌ in advanced digital skills‌ education in Europe.

AEx Smolphone​​

Participants: Martin Quinson,​​​‌ Victorien Elvinger.

• Title:‌
  SmolPhone: un smartphone‌​‌ conscient des limites énergétiques.​​
• Partner Institution(s):
   

  • Inria Magellan​​​‌ team
  • Inria Taran team‌
• Date/Duration:
  2024-2026
• Additionnal info/keywords:‌​‌
  The SmolPhone project is​​ research action exploring potential​​​‌ frugal and low-tech solutions‌ in the domain of‌​‌ mobile computing. Its practical​​ aspects consist in designing​​​‌ a sort of low-tech‌ smartphone offering some services‌​‌ of a classical smartphone​​ with a one-week battery​​​‌ lifetime. The goal of‌ the AEx project is‌​‌ to unlock several technical​​ difficulties by conceiving a​​​‌ software and hardware development‌ workbench enabling future research‌​‌ on the topic of​​ frugal computing and low-tech​​​‌ IT services.

ADT SmartObs‌

Participants: Guillaume Pierre,‌​‌ Matthieu Nicolas.

• Title:​​
  Une plateforme originale pour​​​‌ le monitoring environemental
• Partner‌ Institution(s):
   

  • Inria Magellan team‌​‌
  • Inria Taran team
• Date/Duration:​​
  2024-2025
• Additionnal info/keywords:
  The​​​‌ goal of the SmartObs‌ project is to help‌​‌ develop the SmartSense and​​ LivingFog platforms dedicated to​​​‌ natural environment monitoring, and‌ to enable interdisciplinary collaborations‌​‌ with partners from the​​ environmental sciences domain. The​​​‌ objective is to generate,‌ analyse and transmit smart‌​‌ sensor data (SmartSense) and​​ further process these data​​​‌ in situ (LivingFog) before‌ sending them to a‌​‌ remote cloud.

ANR Dark-Era​​

Participants: Martin Quinson,​​​‌ Ophélie Renaud.

• Title:‌
  Dark-Era: Dataflow Algorithm‌​‌ aRchitecture co-design of SKA​​ pipeline for Exascale RadioAstronomy​​​‌
• Partner Institution(s):
   

  • CentraleSupélec, Laboratoire‌ des Signaux et Systèmes‌​‌ (LSN)
  • INSA Rennes, Institut​​​‌ d'Electronique et de Télécommunication​ de Rennes (IETR)
  • Observatoire​‌ Paris, Galaxies, Etoiles, Physique,​​ Instrumentation
  • Observatoire de la​​​‌ Côte d'Azur Nice, Laboratoire​ J-L. Lagrange
  • ENS Rennes,​‌ Institut de Recherche en​​ Informatique et Systèmes Aléatoires​​​‌ (IRISA)
• Date/Duration:
  2021-2025
• Additionnal​ info/keywords:
  The future Square​‌ Kilometer Array (SKA) radio​​ telescope poses unprecedented challenges​​​‌ to the underlying computational​ system. The instrument is​‌ expected to produce a​​ sustained rate of Terabytes​​​‌ of data per second,​ mandating on-site pre-processing to​‌ reduce the size of​​ data to be transferred.​​​‌ However, the electromagnetic noise​ of a traditional computing​‌ center would hinder the​​ quality of the measurements​​​‌ if located near to​ the instrument. As a​‌ result, the Science Data​​ Processor (SDP) pipeline will​​​‌ only have an energy​ budget of only 1​‌ MWatt to execute a​​ complex algorithm chain estimated​​​‌ at 250 Petaops/s. Because​ of these requirements, the​‌ SDP must be an​​ innovative data-oriented infrastructure running​​​‌ on a disaggregated architecture​ combining standard HPC systems​‌ with dedicated accelerators such​​ as GPU or FPGA.​​​‌ The goal of the​ DarkEra project is to​‌ contribute to the performance​​ assessment both in time​​​‌ and energy of new​ complex scientific algorithms on​‌ not-yet-existing complex computing infrastructures.​​ To that extend, a​​​‌ prototyping tool is developed​ for the prospective profiling​‌ of data-oriented applications during​​ their development.

ANR FACTO​​​‌

Participants: Anne-Cécile Orgerie,​ François Lemercier, Aya​‌ Shehata.

• Title:
  FACTO​​: A Multi-Purpose Wi-Fi​​​‌ Network for a Low-Consumption​ Smart Home
• Partner Institution(s):​‌
   

  • CNRS, IRISA lab
  • University​​ of Lyon 1, LIP​​​‌ lab
  • Orange
  • Fondation Blaise​ Pascal
• Date/Duration:
  2021-2024
• Additionnal​‌ info/keywords:
  The number of​​ smart homes is rapidly​​​‌ expanding worldwide with an​ increasing amount of wireless​‌ IT devices. The diversity​​ of these devices is​​​‌ accompanied by the development​ of multiple wireless protocols​‌ and technologies that aim​​ to connect them. However,​​​‌ these technologies offer overlapping​ capabilities. This overprovisioning is​‌ highly suboptimal from an​​ energy point of view​​​‌ and can be viewed​ as a first barrier​‌ towards sustainable smart homes.​​ Therefore, in the FACTO​​​‌ project, we propose to​ design a multi-purpose network​‌ based on a single​​ optimized technology (namely Wi-Fi),​​​‌ in order to offer​ an energy-efficient, adaptable and​‌ integrated connectivity to all​​ smart home's devices.

ANR​​​‌ EDEN4SG

Participants: Anne-Cécile Orgerie​, Matthieu Silard.​‌

• Title:
  EDEN4SG: Efficient​​ and Dynamic ENergy Management​​​‌ for Large-Scale Smart Grids​
• Partner Institution(s):
   

  • CNRS, IRISA​‌ lab
  • CNRS, SATIE lab​​
  • University of Toulouse, IRIT​​​‌ lab
  • EDF
  • SRD Energies​
  • Orange
• Date/Duration:
  2023-2027
• Additionnal​‌ info/keywords:
  Climate change as​​ well as geopolitical tensions​​​‌ have led a large​ number of countries to​‌ target a massive integration​​ of renewables in their​​​‌ energy mix. This will​ be achieved among others​‌ by increasing the electrification​​ rate of several sectors​​​‌ such as transport. In​ this context, the wide-scale​‌ deployment of electrical vehicles​​ (EVs) represents a challenge​​​‌ as well as an​ opportunity to render more​‌ efficient and affordable the​​ transformation of the current​​​‌ power system into a​ smarter grid. The project​‌ targets to develop methods​​ for the intelligent coordination​​ of large-scale EV fleets​​​‌ and as well to‌ determine the associated cost‌​‌ of information for piloting​​ the required smart grid.​​​‌

IPCEI-CIS E2CC (BPI)

Participants:‌ Anne-Cécile Orgerie, Anudipa‌​‌ Mondal, Hayfa Tayeb​​.

• Title:
  Eco Edge​​​‌ to Cloud Continuum
• Partner‌ Institution(s):
   

  • CNRS, IRISA lab‌​‌
  • Inria, LIG lab
  • University​​ of Toulouse, IRIT lab​​​‌
  • CNRS, LAAS lab
  • ATOS‌ (Bull SAS)
  • Armadillo
  • CGI‌​‌ France
  • NBS
  • Ningaloo
  • Provenrun​​
  • Ryax
• Date/Duration:
  2024-2028
• Additionnal​​​‌ info/keywords:
  The aim of‌ the IPCEI-CIS E2CC project‌​‌ is to provide a​​ standardized end-to-end platform that​​​‌ can be connected natively‌ to any cloud provider,‌​‌ offering the right level​​ of interoperability and service​​​‌ portability and creating a‌ continuum from Edge to‌​‌ Cloud, including cybersecurity, decarbonisation,​​ orchestration and platform functions.​​​‌ This platform will provide‌ hardware and software solutions,‌​‌ exhibiting vertical services (computer​​ vision, decarbonisation services, etc.),​​​‌ cybersecurity services, AI/MLOPs, and‌ Baremetal & Edge infrastructures.‌​‌

CARECloud (PEPR Cloud)

Participants:​​ Anne-Cécile Orgerie, Matteo​​​‌ Chancerel, Pablo Leboulanger‌, Thomas Stavis.‌​‌

• Title:
  CARECLOUD: Understanding,​​ improving, reducing the environmental​​​‌ impacts of Cloud Computing.‌
• Partner Institution(s):
   

  • CNRS, IRISA‌​‌ lab
  • Inria, CRIStAL lab​​
  • CNRS, I3S lab
  • University​​​‌ of Toulouse, IRIT lab‌
  • Inria, LIP lab
  • IMT,‌​‌ LS2N lab
  • IMT, SAMOVAR​​ lab
• Date/Duration:
  2023-2030
• Additionnal​​​‌ info/keywords:
  Cloud computing and‌ its many variations offer‌​‌ users considerable computing and​​ storage capacities. The maturity​​​‌ of virtualization techniques has‌ enabled the emergence of‌​‌ complex virtualized infrastructures, capable​​ of rapidly deploying and​​​‌ reconfiguring virtual and elastic‌ resources in increasingly distributed‌​‌ infrastructures. This resource management,​​ transparent to users, gives​​​‌ the illusion of access‌ to flexible, unlimited and‌​‌ almost immaterial resources. However,​​ the power consumption of​​​‌ these clouds is very‌ real and worrying, as‌​‌ are their overall greenhouse​​ gas (GHG) emissions and​​​‌ the consumption of critical‌ raw materials used in‌​‌ their manufacture. In a​​ context where climate change​​​‌ is becoming more visible‌ and impressive every year,‌​‌ with serious consequences for​​ people and the planet,​​​‌ all sectors (transport, building,‌ agriculture, industry, etc.) must‌​‌ contribute to the effort​​ to reduce GHG emissions.​​​‌ Despite their ability to‌ optimize processes in other‌​‌ sectors (transport, energy, agriculture),​​ clouds are not immune​​​‌ to this observation: the‌ increasing slope of their‌​‌ greenhouse gas emissions must​​ be reversed, otherwise their​​​‌ potential benefits in other‌ sectors will be erased.‌​‌ This is why the​​ CARECloud project (understanding, improving,​​​‌ reducing the environmental impacts‌ of Cloud Computing) aims‌​‌ to drastically reduce the​​ environmental impacts of cloud​​​‌ infrastructures.

CNRS DISTANT

Participants:‌ Guillaume Pierre, Ammar‌​‌ Kazem, Matthieu Nicolas​​.

• Title:
  Automated D​​​‌ata qualIty‌ asSurance for‌​‌ a criTical​​ zone observAtory​​​‌ iNThe‌ Himalayas – Kaligandaki River‌​‌ Nepal
• Partner Institution(s):
   

  • Université​​ de Rennes – Géosciences​​​‌ Rennes, France
  • Université de‌ Rennes – IRISA, France‌​‌
  • IRD, France
  • German Research​​ Centre for Geosciences GFZ,​​​‌ Germany
  • Tribhuvan University, Nepal‌
  • Smartphones4Water Citizen Science NGO,‌​‌ Nepal
• Date/Duration:
  2024-2025
• Additionnal​​ info/keywords:
  Mountain landscapes respond​​​‌ disproportionately quickly to changes‌ in external forcing by‌​‌ tectonics and climate. Erosion​​​‌ and associated extreme events​ are common and present​‌ a continuous threat to​​ local populations. At the​​​‌ same time, mountains act​ as resources and supply​‌  1/3 of the world​​ population with fresh water.​​​‌ Yet mountains are notoriously​ under studied and are​‌ widely underrepresented in global​​ monitoring networks. One particular​​​‌ problem of monitoring such​ landscapes is the data​‌ quality, especially in remote​​ areas where manned support​​​‌ is difficult to maintain.​ The dedicated monitoring area​‌ is the Kaligandaki Catchment​​ in Nepal, spanning an​​​‌ ideal transect across the​ Himalayan Mountain Range from​‌ the edge of the​​ Tibetan Plateau, through the​​​‌ high mountains, to the​ low-lying Gangetic foreland. We​‌ will implement a LivingFog​​ platform computing and data​​​‌ logging system that can​ detect bogus data and​‌ issue alerts to the​​ operators. This system is​​​‌ designed to serve as​ a near real-time data​‌ processing system for early​​ warning of catastrophic events​​​‌ in the future.

NF-JEN​ (PEPR 5G and future​‌ networks)

Participants: Anne-Cécile Orgerie​​, Ibtissem Oueslati.​​​‌

• Title:
  NF-JEN: Just​ Enough Networks
• Partner Institution(s):​‌
   

  • CNRS, IRISA lab
  • CEA,​​ LETI
  • INSA de Lyon,​​​‌ CITI lab
  • Inria, CRIStAL​ lab
  • IMT, IEMN lab​‌
  • CNRS, IETR lab
  • INP​​ Bordeaux, IMS lab
  • IMT,​​​‌ Lab-STICC lab
  • ESIEE -​ Université Gustave Eiffel, LIGM​‌ lab
  • Inria, LIP lab​​
  • IMT, LTCI lab
  • CNRS,​​​‌ XLIM lab
• Date/Duration:
  2023-2027​
• Additionnal info/keywords:
  Communication networks​‌ are often presented as​​ a necessary means of​​​‌ reducing the impact on​ the environment of various​‌ sectors of industry. In​​ practice, the roll-out of​​​‌ new generations of mobile​ broadband networks has required​‌ increased communication resources for​​ wireless access networks. This​​​‌ has proved an effective​ approach in terms of​‌ performance but concerns remain​​ about its energy cost​​​‌ and more generally its​ environmental impacts. Exposure to​‌ electromagnetic fields also remains​​ a cause of concern​​​‌ despite existing protection limits.​ In the JEN (Just​‌ Enough Networks) project, we​​ propose to develop just​​​‌ enough networks: network whose​ dimension, performance, resource usage​‌ and energy consumption are​​ just enough to satisfy​​​‌ users needs. Along with​ designing energy-efficient and sober​‌ networks, we will provide​​ multi-indicators models that could​​​‌ help policy-makers and inform​ the public debate.

STEEL​‌ (PEPR Cloud)

Participants: Cédric​​ Tedeschi, Mathis Valli​​​‌, Minh Thong Le​ Viet.

• Title:
  STEEL​‌: Secure and Efficient​​ Data Storage and Processing​​​‌ on Cloud-based Infrastructures
• Partner​ Institution(s):
   

  • DRIM (CNRS, INSA​‌ Lyon)
  • ERODS (Université Grenoble​​ Alpes)
  • Kerdata (Inria, INSA​​​‌ Rennes)
  • IN2P3 (CNRS)
  • MAGELLAN​ (Université de Rennes, Inria)​‌
  • PROGRESS (Université de Bordeaux,​​ INP)
  • STACK (IMT Atlantique,​​​‌ Inria, LS2N)
  • TeraLab (IMT)​
• Date/Duration:
  7 years
• Additionnal​‌ info/keywords:
  The strong development​​ of cloud computing and​​​‌ its massive adoption for​ the storage of unprecedented​‌ volumes of data in​​ a growing number of​​​‌ domains has brought to​ light major technological challenges.​‌ In this project we​​ will address several of​​​‌ these challenges, organized in​ three research directions:

  • Exploitation​‌ of emerging technologies for​​ efficient storage on cloud​​​‌ infrastructures. We will address​ this challenge through NVRAM-based​‌ distributed performance storage solutions,​​ as close as possible​​ to data production and​​​‌ consumption locations (disaggregation principle)‌ and develop strategies to‌​‌ optimize the trade-off between​​ data consistency and access​​​‌ performance.
  • Efficient storage and‌ processing of data on‌​‌ hybrid, heterogeneous infrastructures within​​ the digital edge-cloud-supercomputer continuum.​​​‌ The execution of hybrid‌ workflows combining simulations, analysis‌​‌ of sensor data flows​​ and machine learning requires​​​‌ storage resources ranging from‌ the edge to cloud‌​‌ infrastructures, and even to​​ supercomputers, which poses challenges​​​‌ for unified data storage‌ and processing.
  • Confidential storage,‌​‌ in connection with the​​ need to store and​​​‌ analyze large volumes of‌ data of strategic interest‌​‌ or of a personal​​ nature.

Taranis (PEPR Cloud)​​​‌

Participants: Shadi Ibrahim,‌ Nikos Parlavantzas, Ahmed‌​‌ Rjiba, Tiago Da​​ Silva Barros.

• Title:​​​‌
  Taranis: Model, Deploy,‌ Orchestrate, and Optimize Cloud‌​‌ Applications and Infrastructure
• Partner​​ Institution(s):
   

  • IMT, France
  • Université​​​‌ Grenoble Alpes, France
  • Université‌ de Rennes, France
  • Inria,‌​‌ France
  • CNRS, France
  • CEA,​​ France
  • ENS Lyon, France​​​‌
  • Université Claude Bernard Lyon‌ 1, France
  • Université de‌​‌ Lille, France
  • INSA Rennes,​​ France
• Date/Duration:
  7 years​​​‌
• Additionnal info/keywords:
  New infrastructures,‌ such as Edge Computing‌​‌ or the Cloud-Edge-IoT computing​​ continuum, complicate the cloud​​​‌ landscape as they add‌ new challenges related to‌​‌ resource diversity and heterogeneity​​ (from small sensors to​​​‌ data centers/HPC, from low‌ power networks to core‌​‌ networks), geographical distribution, as​​ well as increased dynamicity​​​‌ and security needs, all‌ under energy consumption and‌​‌ regulatory constraints. In order​​ to efficiently exploit new​​​‌ infrastructures, we propose a‌ strategy based on a‌​‌ significant abstraction of the​​ application structure description to​​​‌ further automate application and‌ infrastructure management. Thus, it‌​‌ will be possible to​​ globally optimize used resources​​​‌ with respect to multi-criteria‌ objectives (price, deadline, performance,‌​‌ energy, etc.) on both​​ the user side (applications)​​​‌ and the provider side‌ (infrastructures). This abstraction also‌​‌ includes the challenges related​​ to facilitating application reconfiguration​​​‌ and to automatically adapting‌ the use of resources.‌​‌ The Taranis project addresses​​ these issues through four​​​‌ scientific work packages, each‌ focusing on a phase‌​‌ of the application lifecycle:​​ application and infrastructure description​​​‌ models, deployment and reconfiguration,‌ orchestration, and optimization. Members‌​‌ of the Magellan team​​ contribute to four sub-topics​​​‌ including (1) Decentralized, market-based‌ application orchestration for Fog‌​‌ and IoT environments; (2)​​ Realizing and optimizing Serverless​​​‌ Computing in the Edge-Cloud‌ continuum; (3) Orchestrating multi-dimensional‌​‌ resources in the Edge-Cloud​​ continuum; and (4) Resource​​​‌ Provisioning for stream data‌ processing in the Fog.‌​‌

IPCEI-CIS DXP

Participants: Shadi​​ Ibrahim, Cédric Tedeschi​​​‌.

• Title:
  Data Exchange‌
• Partner Institution(s):
   

  • Inria
  • Amadeus‌​‌
• Date/Duration:
  2024-2029
• Additionnal info/keywords:​​
  The aim of the​​​‌ IPCEI-CIS DXP project is‌ to to design and‌​‌ develop an open source​​ management solution for a​​​‌ federated and distributed data‌ exchange platform operating in‌​‌ an open, scalable, and​​ massively distributed Cloud-Edge Continuum.​​​‌

VideoImpact‌ (Labex Cominlabs)

Participants: Anne-Cécile‌​‌ Orgerie, Robin Richard​​, Natacha Lapeyroux.​​​‌

• Title:
  VideoImpact: Model,‌ Deploy, Orchestrate, and Optimize‌​‌ Cloud Applications and Infrastructure​​
• Partner Institution(s):
   

  • Inria
  • CNRS,​​​‌ IRISA
  • Université de Rennes,‌ ARENES
  • INSA de Rennes,‌​‌ IETR
  • Université Catholique de​​​‌ l'Ouest de Nantes, ARENES​
  • IMT Atlantique, LEGO
• Date/Duration:​‌
  2 years
• Additionnal info/keywords:​​
  Recent studies forecast a​​​‌ global warming of 3.1°C​ in 2100 if the​‌ GHG emissions do not​​ decrease. Hence, every part​​​‌ of our society must​ urgently aim sobriety, including​‌ the digital world, that​​ is not intangible, contrary​​​‌ to popular belief. Video​ consumption takes a significant​‌ part among the emissions​​ of the digital world​​​‌ and constitutes a representative​ example of unbounded and​‌ energy-consuming digital system. In​​ that context, a crucial​​​‌ question to tackle is​ how to set limits​‌ to the deployment of​​ a digital system, and​​​‌ for example to video​ delivery systems? This question​‌ is, by nature, lying​​ at the crossroad of​​​‌ many fields (including human​ and social sciences). Interestingly,​‌ many initiatives have recently​​ emerged at the regional​​​‌ level, e.g., the rapprochement​ between the GIS Marousin​‌ and video processing scientists​​ of INSA and IRISA,​​​‌ and set interesting perspectives​ of wide collaborative user​‌ experiments. In that context,​​ the VideoImpact project proposes​​​‌ to answer the following​ questions: In order to​‌ set a sobriety policy,​​ what should we limit​​​‌ in priority? the number​ of hours spent by​‌ a user watching videos?​​ The TV screen size?​​​‌ The video resolutions? The​ deployment of more efficient​‌ digital infrastructure? The VideoImpact​​ project aims at developing​​​‌ i) an environmental footprint​ model for the video​‌ delivery chain to identify​​ the clear levers to​​​‌ sobriety, ii) a solid​ network of industrial and​‌ academic partners of the​​ Rennes' neighborhood around the​​​‌ goal of reducing the​ environmental impact of video​‌ consumption and iii) to​​ launch a concrete experimentation​​​‌ in collaboration with Human​ and Social scientists. The​‌ conclusions will be used​​ in the context of​​​‌ further collaborations with Human​ and Social Scientists to​‌ set real user experiments​​ to assess the feasibility​​​‌ and acceptance of such​ levers.

11.1.1​​ Scientific events: organisation

11.1.2 Scientific events: selection​​

11.1.3 Journal

11.1.4‌​‌ Invited talks

• Guillaume Pierre​​ gave a keynote on​​​‌ “Fog computing, from smart‌ city services to natural‌​‌ environment monitoring” at the​​ 43rd Brazilian Symposium on​​​‌ Computer Networks and Distributed‌ Systems, May 22rd 2025,‌​‌ Natal, Brazil 43.​​
• Anne-Cécile Orgerie gave a​​​‌ keynote on “Carbon and‌ water footprints of computing‌​‌ infrastructures: evaluations and tendencies”​​ at the Infrastructure Workshop,​​​‌ Nantes, France, December 5,‌ 2025.
• Anne-Cécile Orgerie gave‌​‌ a keynote on “Carbon​​ Footprint Allocation Models in​​​‌ Distributed Systems” at UCC‌ (IEEE/ACM International Conference on‌​‌ Utility and Cloud Computing),​​ Nantes, France, December 4,​​​‌ 2025.
• Anne-Cécile Orgerie gave‌ a keynote on “Greening‌​‌ ICT: challenges and research​​ directions in the field​​​‌ of distributed systems towards‌ sustainability” at the Green‌​‌ ICT and ICT for​​​‌ Green workshop co-located with​ Informatics Europe Summit (ECSS),​‌ Rennes, France, October 29,​​ 2025.
• Anne-Cécile Orgerie gave​​​‌ a keynote on “Energy​ Consumption and Environmental Impact​‌ of Distributed Systems” at​​ ISPDC (IEEE International Symposium​​​‌ on Parallel and Distributed​ Computing), Rennes, France, July​‌ 10, 2025.
• Shadi Ibrahim​​ gave an invited talk​​​‌ on “Scalable and Efficient​ Big Data Processing in​‌ Clouds: Addressing Performance Variability”'​​ at the Fifth Workshop​​​‌ on Challenges and Opportunities​ of Efficient and Performant​‌ Storage Systems (CHEOPS'25), Rotterdam,​​ Netherlands, March 31 2025​​​‌

11.1.5 Leadership within the​ scientific community

• Anne-Cécile Orgerie​‌ is the director of​​ the CNRS research and​​​‌ service group on ICT​ environmental impact (GDRS EcoInfo).​‌
• Anne-Cécile Orgerie is the​​ chief scientist for the​​​‌ Rennes site of Grid'5000​ and co-responsible for the​‌ Rennes site of SLICES-FR.​​

11.1.6 Scientific expertise

• Anne-Cécile​​​‌ Orgerie was an HCERES​ expert for the evaluation​‌ of a laboratory in​​ 2025.
• Anne-Cécile Orgerie was​​​‌ a reviewer for a​ research proposal for the​‌ FWF Austrian Science Fund​​ in 2025.
• Anne-Cécile Orgerie​​​‌ was a reviewer for​ a research proposal for​‌ the Norvegian Velux Foundation​​ in 2025.
• Shadi Ibrahim​​​‌ was a member of​ the ACM Heidelberg Laureate​‌ Forum (HLF) 2025 Young​​ Researcher Selection Committee.

11.1.7​​​‌ Research administration

• Anne-Cécile Orgerie​ is an officer (chargée​‌ de mission) for the​​ IRISA cross-cutting axis on​​​‌ Green IT.
• Anne-Cécile Orgerie​ is a member of​‌ the steering committee of​​ the CNRS GDR RSD.​​​‌
• Shadi Ibrahim co-organizies the​ SCI-Rennes seminar: a monthly​‌ series of scientific seminars​​ for all staff at​​​‌ Inria research centre at​ Rennes University.
• Guillaume Pierre​‌ is a member of​​ Inria Rennes center's “bureau​​​‌ du comité des projets”​

11.2.1​​​‌ Teaching

• Bachelor: Marin Bertier​ , Programmation en Java,​‌ L1, INSA Rennes.
• Bachelor:​​ Marin Bertier , Langage​​​‌ C, L3, INSA Rennes.​
• Bachelor: Marin Bertier ,​‌ Réseaux, L3, INSA Rennes.​​
• Bachelor: François Lemercier ,​​​‌ Système Cloud et Réseaux,​ L3 Informatique, Univ. Rennes.​‌
• Bachelor: François Lemercier ,​​ Mineure Cloud et Réseaux,​​​‌ L3 Informatique, Univ. Rennes.​
• Bachelor: Anne-Cécile Orgerie ,​‌ Impacts environnementaux du numérique,​​ classe CPES, Lycée Chateaubriand,​​​‌ Rennes.
• Bachelor: Nikos Parlavantzas​ , Databases, L2, INSA​‌ Rennes.
• Bachelor: Guillaume Pierre​​ , Systèmes Informatiques, L3​​​‌ MIAGE, Univ. Rennes.
• Bachelor:​ Guillaume Pierre , Systèmes​‌ d'exploitation, L3 Informatique, Univ.​​ Rennes.
• Bachelor: Martin Quinson​​​‌ , Architecture et Systèmes,​ L3 Informatique, ENS Rennes.​‌
• Bachelor: Martin Quinson ,​​ Pédagogie, L3 Informatique, ENS​​​‌ Rennes.
• Bachelor: Cédric Tedeschi​ , Systèmes, Cloud et​‌ Réseaux, L3 Informatique, Univ.​​ Rennes.
• Master: Marin Bertier​​​‌ , Systèmes d'exploitation, M1,​ INSA Rennes.
• Master: Shadi​‌ Ibrahim , Hadoop and​​ Cloud Computing, 36hETD, M2​​​‌ : Statistics for Smart​ Data, ENSAI, Bruz.
• Master:​‌ Shadi Ibrahim , NoSQL​​ and Cloud technologies, 45hETD,​​​‌ M2, ENSAI, Bruz.
• Master:​ François Lemercier , Cloud,​‌ M1 RSSI, Univ. Rennes.​​
• Master: François Lemercier ,​​​‌ Network Administration and new​ technologies, M1 CNI, Univ.​‌ Rennes.
• Master: François Lemercier​​ , Advanced Network Infrastructure,​​​‌ M2 Cloud et Réseaux,​ Univ. Rennes.
• Master: Anne-Cécile​‌ Orgerie , Energy efficiency​​ in distributed systems, TelecomSudParis,​​ Palaiseau.
• Master: Anne-Cécile Orgerie​​​‌ , Consommation énergétique des‌ ordinateurs, formation for computer‌​‌ science preparatory class teachers,​​ CIRM, Marseille.
• Master: Nikos​​​‌ Parlavantzas , 4th-year Project,‌ M1, INSA Rennes.
• Master:‌​‌ Nikos Parlavantzas , Clouds,​​ M1, INSA Rennes.
• Master:​​​‌ Nikos Parlavantzas , IoT,‌ M2, INSA Rennes.
• Master:‌​‌ Nikos Parlavantzas , Smart​​ city services, M2 CNI,​​​‌ Univ. Rennes.
• Master: Nikos‌ Parlavantzas , NoSQL, M2,‌​‌ Master for Smart Data​​ Science, ENSAI.
• Master: Nikos​​​‌ Parlavantzas , SQL, M2,‌ Master for Smart Data‌​‌ Science, ENSAI.
• Master: Guillaume​​ Pierre , Distributed Systems,​​​‌ M1 CR&CNI, Univ. Rennes.‌
• Master: Guillaume Pierre ,‌​‌ Systèmes répartis, ENSAI.
• Master:​​ Guillaume Pierre , Services​​​‌ for Cloud technology, M1‌ CR&CNI, Univ. Rennes.
• Master:‌​‌ Guillaume Pierre , Advanced​​ Cloud Infrastructures, M2 CR&CNI&SIF,​​​‌ Univ. Rennes.
• Master: Martin‌ Quinson , C++ system‌​‌ programming, ENS Rennes.
• Master:​​ Martin Quinson , HPC​​​‌ programming, ENS Rennes.
• Master:‌ Martin Quinson , Préparation‌​‌ à l'Agrégation d'Informatique (Networking,​​ 20h ETD, Operating systems,​​​‌ 20h ETD), ENS Rennes.‌
• Master: Martin Quinson ,‌​‌ Scientific Outreach, M2, 30​​ hEDT, ENS Rennes.
• Master:​​​‌ Cédric Tedeschi , Concurrences‌ et Système d'Exploitation, M1‌​‌ Informatique, Univ. Rennes.
• Master:​​ Cédric Tedeschi , Concurrence​​​‌ et coopération dans les‌ systèmes et les réseaux,‌​‌ M1 MIAGE. Univ. Rennes.​​
• Master: Cédric Tedeschi ,​​​‌ Advanced Cloud Infrastructures, M2‌ Informatiquie, Univ. Rennes.
• Master:‌​‌ Cédric Tedeschi , Parallel​​ Programming with Python, ENSAI​​​‌

11.2.2 Supervision

• PhD defended:‌ Khaled Arsalane , “Scalable‌​‌ Data Stream Processing in​​ Heterogeneous Environments” defended on​​​‌ December 15th 2025 at‌ the Université de Rennes‌​‌ 36.
• PhD defended:​​ Léo Cosseron , “Time-accurate​​​‌ network simulation interconnecting virtual‌ machines with hardware virtualization‌​‌ towards stealth analysis” defended​​ on December 8th 2025​​​‌ at ENS Rennes 2‌.
• PhD in progress:‌​‌ Maxime Agusti , “Observation​​ of baremetal co-location platforms,​​​‌ models and catalog proposal‌ to reduce energy consumption”,‌​‌ started in 2021, supervised​​ by Eddy Caron ,​​​‌ Laurent Lefèvre and Anne-Cécile‌ Orgerie .
• PhD in‌​‌ progress: Matteo Chancerel ,​​ “Optimization of a fully​​​‌ distributed Fog powered by‌ renewable energy sources”, started‌​‌ in 2024, supervised by​​ Anne-Cécile Orgerie .
• PhD​​​‌ in progress: Ammar Kazem‌ , “Systèmes de traitement‌​‌ de données in-natura pour​​ l’observation environnementale sous contrainte​​​‌ énergétique”', started in 2023,‌ supervised by Guillaume Pierre‌​‌ and Laurent Longuevergne .​​
• PhD in progress: Govind​​​‌ Kovilkkatt Panickerveetil , “Traitement‌ de flux de données‌​‌ économe en énergie”', started​​ in 2023, supervised by​​​‌ Guillaume Pierre and Romain‌ Rouvoy .
• PhD in‌​‌ progress: Haraesh Jayasethu Ramachandran​​ , “Latency-driven container network​​​‌ optimization in edge industrial‌ IoT,” started in 2025,‌​‌ supervised by François Lemercier​​ and Guillaume Pierre .​​​‌
• PhD in progress: Pablo‌ Leboulanger , “Minimalist cloud,‌​‌ low on energy, hardware​​ and software resources”, started​​​‌ in 2024, supervised by‌ Anne-Cécile Orgerie .
• PhD‌​‌ in progress: Robin Richard​​ , “Limits for the​​​‌ global video consumption”, started‌ in 2025, supervised by‌​‌ Thomas Maugey and Anne-Cécile​​ Orgerie .
• PhD in​​​‌ progress: Matthieu Silard ,‌ “Co-optimization of electrical and‌​‌ communication networks”, started in​​ 2023, supervised by Anne-Cécile​​​‌ Orgerie , Nicolas Montavont‌ and Georgios Papadopoulos .‌​‌
• PhD in progress: Thomas​​​‌ Stavis , “Replay of​ environmental leverages in Cloud​‌ infrastructures”, started in 2024,​​ supervised by Laurent Lefèvre​​​‌ and Anne-Cécile Orgerie .​
• PhD in progress: Quentin​‌ Acher , “Management of​​ mutable data over P2P​​​‌ storage”, started in 2023,​ supervised by Shadi Ibrahim​‌ and Claudia-Lavinia Ignat .​​
• PhD in progress: Mohammad​​​‌ Rizk , “Reliable and​ cost-efficient data placement and​‌ repair in P2P storage​​ over immutable data”, started​​​‌ in November 2023, supervised​ by Shadi Ibrahim ,​‌ Thomas Lambert , and​​ Guillaume Pierre .
• PhD​​​‌ in progress: Marc Tranzer​ , “Energy efficient data​‌ management: Data reduction and​​ protection meet performance and​​​‌ energy”, started in January​ 2024, supervised by Shadi​‌ Ibrahim and Guillaume Pierre​​ .
• PhD in progress:​​​‌ Volodia Parol-Guarino , “Flexible​ resource allocation for FaaS​‌ applications in the fog”,​​ started in October 2022,​​​‌ supervised by Nikolaos Parlavantzas​ .
• PhD in progress:​‌ Mohamed Cherif Zouaoui Latreche​​ ,“Balancing Performance and Sustainability​​​‌ for FaaS in the​ Fog”, started in October​‌ 2023, supervised by Nikolaos​​ Parlavantzas and Hector Duran-Limon.​​​‌
• PhD in progress: Ahmed​ Rjiba ,“Decentralised, market-based application​‌ orchestration in Fog and​​ IoT environments”, started in​​​‌ November 2025, supervised by​ Nikos Parlavantzas and Remous-Aris​‌ Koutsiamanis .
• PhD in​​ progress: Stella Lafortune Tchoutcha​​​‌ Sandjong ,“Energy-Efficient FaaS for​ Low-Power Edge and IoT​‌ Deployments”, started in December​​ 2025, supervised by Barbe​​​‌ Thystere Mvondo Djob and​ Nikolaos Parlavantzas .
• PhD​‌ in progress: Mathieu Laurent​​ on Efficient verification of​​​‌ asynchronous distributed systems, started​ in October 2023, supervised​‌ by Thierry Jéron (Devine)​​ and Martin Quinson .​​​‌
• PhD in progress: Minh​ Thong Le Viet on​‌ “Towards a programmable autonomic​​ platform for decentralized learning”,​​​‌ started in October 2025,​ supervised by Cédric Tedeschi​‌ .

11.2.3 Juries

• Anne-Cécile​​ Orgerie was chairperson at​​​‌ the PhD defense of​ Robin Boezennec (Université de​‌ Rennes), December 10th 2025.​​
• Anne-Cécile Orgerie was examinor​​​‌ at the PhD defense​ of Roblex Nana Tchakouté​‌ (Mines de Paris), December​​ 5th 2025.
• Anne-Cécile Orgerie​​​‌ was examinor at the​ PhD defense of Killian​‌ Castillon du Perron (Université​​ de Côte d'Azur), November​​​‌ 21st 2025.
• Anne-Cécile Orgerie​ was examinor at the​‌ PhD defense of Tiago​​ da Silva Barros (Université​​​‌ de Côte d'Azur), November​ 3rd 2025.
• Anne-Cécile Orgerie​‌ was chairperson at the​​ PhD defense of Meriem​​​‌ Ghali (ENS de Lyon),​ October 16th 2025.
• Anne-Cécile​‌ Orgerie was examinor at​​ the PhD defense of​​​‌ Meven Mognol (Université de​ Rennes), July 2nd 2025.​‌
• Anne-Cécile Orgerie was opponent​​ and examinor at the​​​‌ PhD defense of Antoine​ Omond (Artic University of​‌ Norway, Tromso and Université​​ de Nantes), May 23rd​​​‌ 2025.
• Shadi Ibrahim was​ examinor at the PhD​‌ defense of Mazen Ezzeddine​​ (Université Côte d'Azur), September​​​‌ 16th 2025.
• Martin Quinson​ was chairperson at the​‌ PhD defense of Khaled​​ Arsalane (Université de Rennes),​​​‌ December 15th 2025.
• Guillaume​ Pierre was reviewer at​‌ the PhD defense of​​ Aymeric Agon (Sorbonne Université),​​​‌ September 15th 2025.

11.3.1 Productions (articles,​‌ videos, podcasts, serious games,​​ ...)

• Guillaume Pierre gave​​​‌ an interview to Le​ Nouvel Observateur which was​‌ mentioned in article “​​Que dit la panne​​ d’AWS, le cloud d’Amazon,​​​‌ de notre dépendance aux‌ géants du numérique ?‌​‌” on October 21st​​ 2025.
• Guillaume Pierre gave​​​‌ an interview to Brief.Science‌ which was mentioned in‌​‌ article “Le cloud,​​ moteur invisible de nos​​​‌ vies connectées” on‌ October 31st 2025.
• Anne-Cécile‌​‌ Orgerie gave an interview​​ to Les Echos which​​​‌ was mentioned in the‌ article “Environnement :‌​‌ faut-il moraliser l'usage du​​ numérique” on February​​​‌ 10th 2025.
• Anne-Cécile Orgerie‌ gave an interview to‌​‌ Mediapart which was mentioned​​ in the article “​​​‌Les déchets du numérique,‌ un fléau pour la‌​‌ ville canadienne de Rouyn-Noranda​​” on February 10th​​​‌ 2025.
• Anne-Cécile Orgerie gave‌ an interview to Le‌​‌ Parisien which was mentioned​​ in the article “​​​‌Dans la course à‌ l'IA, des machines françaises‌​‌ parmi les plus puissantes​​ au monde” on​​​‌ February 23rd 2025.

11.3.2‌ Participation in Live events‌​‌

• Anne-Cécile Orgerie proposed a​​ session at the popularization​​​‌ day on “les décodeuses”‌ with 550 students from‌​‌ 15 high schools.

11.3.3​​ Others science outreach relevant​​​‌ activities

• François Lemercier proposed‌ “Clusterize” in collaboration with‌​‌ artist Loïg Nguyen in​​ the context of the​​​‌ IRISA program “1‌ Artist 1 Scientist.”‌​‌

International journals​​​‌

• 3 articleH.Henri‌ Casanova, A.Arnaud‌​‌ Giersch, A.Arnaud​​ Legrand, M.Martin​​​‌ Quinson and F.Frédéric‌ Suter. Lowering entry‌​‌ barriers to developing custom​​ simulators of distributed applications​​​‌ and platforms with SimGrid‌.Parallel Computing123‌​‌March 2025, 103125​​HALDOIback to​​​‌ text
• 4 articleA.‌Ammar Kazem, G.‌​‌Guillaume Pierre and L.​​Laurent Longuevergne. Enhancing​​​‌ Environmental Observatories with Fog‌ Computing.Frontiers in‌​‌ Environmental Science13April​​ 2025, 1-15HAL​​​‌DOIback to text‌
• 5 articleS.Simon‌​‌ Lambert, V.Vladimir​​ Ostapenco, L.Laurent​​​‌ Lefèvre, E.Eddy‌ Caron, A.-C.Anne-Cécile‌​‌ Orgerie, B.Benjamin​​ Fichel and R.Rémi​​​‌ Grivel. Revisiting virtual‌ machine consolidation to save‌​‌ resources and energy in​​ heterogeneous production cloud infrastructures​​​‌.International Journal of‌ High Performance Computing Applications‌​‌December 2025HALDOI​​back to text
• 6​​​‌ articleJ. C.Jann‌ Camilo Sánchez Huertas,‌​‌ G.Greta Vallero,​​ L.Loutfi Nuaymi and​​​‌ A.-C.Anne-Cécile Orgerie.‌ Review and classification of‌​‌ the use of SMs​​ for energy saving in​​​‌ next-generation radio access networks‌.Computer Networks273‌​‌December 2025, 111745​​​‌HALDOIback to​ text
• 7 articleA.​‌Antonin Voyez, T.​​Tristan Allard, G.​​​‌Gildas Avoine, P.​Pierre Cauchois, E.​‌Elisa Fromont and M.​​Matthieu Simonin. The​​​‌ Privacy Cost of Fine-Grained​ Electrical Consumption Data.​‌Scientific Reports1517391​​May 2025, 1-8​​​‌HALDOIback to​ text
• 8 articleH.​‌Hansheng Zhang, S.​​Song Wu, H.​​​‌Hao Fan, Z.​Zhuo Huang, W.​‌Weibin Xue, C.​​Chen Yu, S.​​​‌Shadi Ibrahim and H.​Hai Jin. KubeSPT:​‌ Stateful Pod Teleportation for​​ Service Resilience With Live​​​‌ Migration.IEEE Transactions​ on Services Computing18​‌3May 2025,​​ 1500-1514HALDOIback​​​‌ to text

National journals​

• 9 articleX.Xavier​‌ Arnauld de Sartre,​​ W.Wurtz Frederic,​​​‌ C.Christophe Bouneau,​ F.Freddy Bouchet,​‌ G.Gilles Debizet,​​ L.Louis Fontenelle (de)​​​‌, L.Lise Desvallées​, M.Marie Forget​‌, N.Nathalie Rodet-Kroichvili​​, S.Sandrine Mathy​​​‌, A.-C.Anne-Cécile Orgerie​, D.Dominique Pecaud​‌, J.Jacques Py​​, O.Olivier Vidal​​​‌ and S.Sébastien Velut​. Pour une science​‌ de la Transition Énergétique​​ au Service de la​​​‌ Société: Rapport du groupe​ interdisciplinaire ARPEGES CNRS pour​‌ la Transition énergétique.​​Natures Sciences Sociétés33​​​‌32025, 1-10​HALback to text​‌

Invited conferences

• 10 inproceedings​​B.Baptiste Jonglez,​​​‌ M.Matthieu Simonin,​ J.Jolan Philippe and​‌ S. M.Sidi Mohammed​​ Kaddour. Multi-provider capabilities​​​‌ in EnOSlib: driving distributed​ system experiments on the​‌ edge-to-cloud continuum.Springer​​ LNCS-IFIP, Lecture Notes in​​​‌ Computer Science (LNCS)DAIS​ 2025: 25th International Conference​‌ on Distributed Applications and​​ Interoperable Systems15730Lille,​​​‌ FranceSpringer2025,​ 25-42HALDOIback​‌ to text

International peer-reviewed​​ conferences

• 11 inproceedingsM.​​​‌Maxime Agusti, E.​Eddy Caron, B.​‌Benjamin Fichel, L.​​Laurent Lefèvre, O.​​​‌Olivier Nicol and A.-C.​Anne-Cécile Orgerie. PPEM-BM:​‌ Portable Power Estimation Methodology​​ for Bare Metal Servers​​​‌.ICPADS 2025 -​ 31st IEEE International Conference​‌ on Parallel and Distributed​​ SystemsICPADS 2025 -​​​‌ 31st IEEE International Conference​ on Parallel and Distributed​‌ SystemsHefei, ChinaIEEE​​2025, 1-8HAL​​​‌back to text
• 12​ inproceedingsK.Khaled Arsalane​‌ and G.Guillaume Pierre​​. Data Stream Processing​​​‌ Effectiveness in Heterogeneous Computing​ Environments.Proceedings of​‌ the 41st ACM/SIGAPP Symposium​​ On Applied ComputingSAC​​​‌ 2026 - 41st ACM/SIGAPP​ Symposium On Applied Computing​‌Tessaloniki, GreeceACM2026​​HALback to text​​​‌
• 13 inproceedingsM.Matteo​ Chancerel and A.-C.Anne-Cécile​‌ Orgerie. Life-cycle carbon​​ assessment of by-design intermittent​​​‌ Fog services fully powered​ by renewable energy sources​‌.UCC 2025 -​​ 18th IEEE/ACM International Conference​​​‌ on Utility and Cloud​ ComputingUCC 2025 -​‌ 18th IEEE/ACM International Conference​​ on Utility and Cloud​​​‌ ComputingNantes, FranceIEEE​2025, 1-10HAL​‌back to text
• 14​​ inproceedingsL.Léo Cosseron​​​‌, L.Louis Rilling​ and M.Martin Quinson​‌. Mitigation of the​​ impact of Virtual Machine​​ Introspection Pauses on Multi-core​​​‌ Virtual Machines.Lecture‌ Notes in Computer Science‌​‌ (LNCS)HS3 2025 -​​ 1st Workshop on Hardware-Supported​​​‌ Software SecurityToulouse, France‌2025, 1-20HAL‌​‌back to text
• 15​​ inproceedingsC.Camille Coti​​​‌ and M.Martin Quinson‌. Simulating a communication‌​‌ middleware for one-sided and​​ collective operations.ISPCS​​​‌ 2025 - International IEEE‌ Symposium on Precision Clock‌​‌ Synchronization for Measurement, Control,​​ & CommunicationOttawa, Canada​​​‌2025, 1-6HAL‌back to text
• 16‌​‌ inproceedingsH.Hao Fan​​, Z.Zhuo Huang​​​‌, S.Shadi Ibrahim‌, L.Lin Gu‌​‌ and S.Song Wu​​. EDDE: Container Deployment​​​‌ Framework Beyond the Cloud‌.SC 2025 -‌​‌ International Conference for High​​ Performance Computing, Networking, Storage​​​‌ and AnalysisSt. Louis‌ MO USA, United States‌​‌ACM2025, 570-585​​HALDOIback to​​​‌ text
• 17 inproceedingsG.‌Gaël Guennebaud, A.-L.‌​‌Anne-Laure Ligozat, A.-C.​​Anne-Cécile Orgerie and M.​​​‌Matthieu Simonin. Evaluating‌ and Reporting the Carbon‌​‌ Footprint of Shared Computing​​ Platforms: Choices and Limits​​​‌.ISPDC 2025 -‌ 24th IEEE International Symposium‌​‌ on Parallel and Distributed​​ ComputingRennes, FranceIEEE​​​‌2025, 1-7HAL‌back to text
• 18‌​‌ inproceedingsS.Shadi Ibrahim​​ and J.Jad Darrous​​​‌. Erasure Coding Aware‌ Block Placement for Data-Intensive‌​‌ Applications.CHEOPS 2025​​ - 5th Workshop on​​​‌ Challenges and Opportunities of‌ Efficient and Performant Storage‌​‌ SystemsRotterdam Netherlands, Netherlands​​ACMMarch 2025,​​​‌ 15-22HALDOIback‌ to text
• 19 inproceedings‌​‌P.Pierre Jacquet,​​ M.Maxime Agusti,​​​‌ E.Eddy Caron,‌ C.Camille Coti,‌​‌ M.Marcos Dias de​​ Assuncao, L.Laurent​​​‌ Lefèvre and A.-C.Anne-Cécile‌ Orgerie. Untangling GPU‌​‌ Power Consumption: Job-Level Inference​​ in Cloud Shared Settings​​​‌.EUROSYS 2026 -‌ European Conference on Computer‌​‌ SystemsEdinbourg, Ecosse, United​​ Kingdom2026HALDOI​​​‌back to text
• 20‌ inproceedingsH.Haraesh Jr‌​‌, A.Ammar Kazem​​, M.Matthieu Nicolas​​​‌, F.François Lemercier‌, G.Guillaume Pierre‌​‌ and C.Christoff Andermann​​. LivingBench: an IoT/Edge​​​‌ Platform Benchmark Based on‌ an Environmental Observation Use-Case‌​‌.ICIN 2026 -​​ 29th Conference on Innovation​​​‌ in Clouds, Internet and‌ NetworksAthens, Greece2026‌​‌HALback to text​​
• 21 inproceedingsA.Ammar​​​‌ Kazem, G.Guillaume‌ Pierre and L.Laurent‌​‌ Longuevergne. Demo: Energy-aware​​ Dynamic Dimensioning of IoT​​​‌ Edge Clusters for Natural‌ Environment Observation.IC2E‌​‌ 2025 - 13th IEEE​​ International Conference on Cloud​​​‌ EngineeringRennes, FranceIEEE‌September 2025, 1-3‌​‌HALback to text​​
• 22 inproceedingsM.Mathieu​​​‌ Laurent, T.Thierry‌ Jéron and M.Martin‌​‌ Quinson. Towards Efficient​​ Verification of Parallel Applications​​​‌ with Mc SimGrid.‌DisCoTec 2025 - 20th‌​‌ International Federated Conference on​​ Distributed Computing TechniquesFORTE​​​‌ 2025 - 45th International‌ Conference on Formal Techniques‌​‌ for Distributed Objects, Components,​​ and SystemsLecture Notes​​​‌ in Computer ScienceLille,‌ FranceSpringer2025,‌​‌ 1-21HALback to​​ text
• 23 inproceedingsP.​​​‌Pablo Leboulanger and A.-C.‌Anne-Cécile Orgerie. Power‌​‌ limits in data centers:​​​‌ what can we expect​ from improving energy efficiency​‌ and refreshing servers?IC2E​​ 2025 - 13th IEEE​​​‌ International Conference on Cloud​ EngineeringIC2E 2025 -​‌ 13th IEEE International Conference​​ on Cloud EngineeringRennes,​​​‌ France2025, 1-12​HALback to text​‌
• 24 inproceedingsF.François​​ Lemercier and A.-C.Anne-Cécile​​​‌ Orgerie. Towards an​ energy-efficient Wi-Fi: An experimental​‌ study on recent standards​​ power consumption.WCNC​​​‌ 2025 - IEEE Wireless​ Communications and Networking Conference​‌Milan, ItalyIEEE2025​​, 01-06HALDOI​​​‌back to text
• 25​ inproceedingsD.Djob Mvondo​‌, B.Boris Djomgwe​​ Teabe and N.Nikos​​​‌ Parlavantzas. Efficient Memory​ Usage For Edge FaaS​‌ Platforms.PerCom 2026​​ - 24th IEEE International​​​‌ Conference on Pervasive Computing​ and CommunicationsPise, Italy​‌2026HALback to​​ text
• 26 inproceedingsV.​​​‌Volodia Parol-Guarino and N.​Nikos Parlavantzas. Auction-based​‌ Placement of Function Chains​​ in the Fog at​​​‌ Scale.Europar 2025​ - 31st International European​‌ Conference on Parallel and​​ Distributed ComputingDresden, Germany​​​‌2025, 1-14HAL​back to text
• 27​‌ inproceedingsO.Ophélie Renaud​​, N.Nicolas Gac​​​‌, F.François Orieux​ and C.Cedric Dumez-Viou​‌. SimSDP, a Rapid​​ Prototyping tool for Radio​​​‌ Astronomy: From NenuFAR Experiments​ to SKAO-Scale Simulation.​‌SiPS 2025 - 38th​​ IEEE Workshop on Signal​​​‌ Processing SystemsHong Kong,​ ChinaIEEE2025,​‌ 1-5HALback to​​ text
• 28 inproceedingsM.​​​‌Matthieu Silard, A.-C.​Anne-Cécile Orgerie, N.​‌Nicolas Montavont and G.​​Georgios Papadopoulos. Dynamic​​​‌ Incentive Signaling through Advanced​ Metering Infrastructure Networks.​‌SmartGridComm 2025 - IEEE​​ International Conference on Communications,​​​‌ Control, and Computing Technologies​ for Smart GridsToronto,​‌ Canada2025, 1-5​​HALback to text​​​‌
• 29 inproceedingsM.Matthieu​ Silard, G.Georgios​‌ Papadopoulos, A.-C.Anne-Cécile​​ Orgerie and N.Nicolas​​​‌ Montavont. Demo: A​ Visualization Platform for Smart​‌ Grid Network.CSCN​​ 2025 - IEEE Conference​​​‌ on Standards for Communications​ and NetworkingBologne, Italy​‌IEEE2025, 1-3​​HALback to text​​​‌

National peer-reviewed Conferences

• 30​ inproceedingsT.Thomas Maugey​‌, D.Daniel Menard​​, A.-C.Anne-Cécile Orgerie​​​‌ and R.Robin Richard​. Diffusion vidéo :​‌ dépasser l'objectif d'efficacité et​​ viser une sobriété choisie​​​‌.GRETSI - XXXe​ Colloque Francophone de Traitement​‌ du Signal et des​​ ImagesGRETSI 2025 -​​​‌ XXXe Colloque Francophone de​ Traitement du Signal et​‌ des ImagesStrasbourg, France​​2025, 1-4HAL​​​‌back to text

Conferences​ without proceedings

• 31 inproceedings​‌J.-T.Jean-Thomas Acquaviva,​​ J.Jalil Boukhobza,​​​‌ P.Philippe Deniel,​ S.Shadi Ibrahim,​‌ P.Philippe Raipin-Parvédy and​​ F.François Trahay.​​​‌ Minutes from the 9​ th edition of the​‌ Performance and Scalability of​​ Storage Systems workshop (Per3S),​​​‌ 23rd May 2025, "Maison​ des Mines et des​‌ Ponts", Paris.9th​​ edition of the workshop​​​‌ Performance and Scalability of​ Storage Systems (Per3S)Paris,​‌ France2025, 1-2​​HALback to text​​​‌
• 32 inproceedingsN.Nicolas​ Gac and O.Ophélie​‌ Renaud. Dark-era :​​ Dataflow Algorithm aRchitecture co-design​​ of SKA pipeline for​​​‌ Exascale Radio Astronomy.‌SaLF 2025 - 11th‌​‌ Annual conference on Science​​ At Low FrequenciesOrléans,​​​‌ France2025HALback‌ to text
• 33 inproceedings‌​‌O.Ophélie Renaud,​​ S.Sunrise Wang,​​​‌ N.Nicolas Gac and‌ F.François Orieux.‌​‌ Extension du prototypage rapide​​ pour les pipelines d’imagerie​​​‌ radioastronomique : Simulation avancée‌ multi-nœud HPC.GRETSI‌​‌ 2025 - 30e Colloque​​ sur le traitement du​​​‌ signal et des images‌Strasbourg, France2025,‌​‌ 1-4HALback to​​ text

Scientific book chapters​​​‌

• 34 inbookW. E.‌Wedan Emmanuel Gnibga,‌​‌ A.Anne Blavette and​​ A.-C.Anne-Cécile Orgerie.​​​‌ Managing Distributed Cloud Infrastructures‌ for Sustainability.Scheduling‌​‌ Variable Capacity Resources for​​ SustainabilityCRC Press2026​​​‌, 83-124HALback‌ to text

Edition (books,‌​‌ proceedings, special issue of​​ a journal)

• 35 periodical​​​‌Serverless Computing.IEEE‌ Internet Computing286‌​‌January 2025, 5-7​​HALDOIback to​​​‌ text

Doctoral dissertations and‌ habilitation theses

• 36 thesis‌​‌K.Khaled Arsalane.​​ Scalable Data Stream Processing​​​‌ in Heterogeneous Environments.‌Université de rennesDecember‌​‌ 2025HALback to​​ textback to text​​​‌

Reports & preprints

• 37‌ reportS.Sylvain Bouveret‌​‌, A.Aurélie Bugeau​​, E.Emmanuelle Frenoux​​​‌, J.Julien Lefevre‌, L.Laurent Lefèvre‌​‌, A.-L.Anne-Laure Ligozat​​, P.Philippe Marquet​​​‌, A.-C.Anne-Cécile Orgerie‌ and D.Denis Trystram‌​‌. Quiz sur les​​ impacts environnementaux du numérique​​​‌.EcoInfoFebruary 2025‌, 1-5HALback‌​‌ to text
• 38 report​​K.Karin Dassas,​​​‌ C.Cyrille Bonamy,‌ B.Bruno Bzeznik,‌​‌ R.Romaric David,​​ E.Emmanuelle Frenoux,​​​‌ G.Gaël Guennebaud,‌ A.-L.Anne-Laure Ligozat,‌​‌ D.Didier Mallarino and​​ A.-C.Anne-Cécile Orgerie.​​​‌ Estimer l'impact carbone des‌ activités numériques de l'Observatoire‌​‌ de Paris.EcoInfo​​February 2025, 1-47​​​‌HALback to text‌
• 39 reportM.Matthieu‌​‌ Simonin. Bilan Carbone​​ du centre de Inria/IRISA​​​‌ de Rennes pour l'année‌ 2024. Les déplacements professionels:‌​‌ dynamiques et disparités entre​​ les équipes du centre​​​‌ Inria de Rennes.: Avec‌ une étude sur les‌​‌ émissions liées aux déplacements​​ professionnels : dynamiques et​​​‌ disparités entre équipes du‌ centre Inria de Rennes‌​‌.Inria Centre at​​ the University of Rennes,​​​‌ France; Irisa Institut de‌ Recherche en Informatique et‌​‌ Systèmes AléatoiresSeptember 2025​​, 1-13HALback​​​‌ to text

Other scientific‌ publications

• 40 miscB.‌​‌Bran Knowles, V.​​ L.Vicki L Hanson​​​‌, C.Christoph Becker‌, M.Mike Berners-Lee‌​‌, A. A.Andrew​​ A Chien, B.​​​‌Benoit Combemale, V.‌Vlad Coroamă, K.‌​‌Koen de Bosschere,​​ Y.Yi Ding,​​​‌ A.Adrian Friday,‌ B.Boris Gamazaychikov,‌​‌ L.Lynda Hardman,​​ S.Simon Hinterholzer,​​​‌ M.Mattias Höjer,‌ L.Lynn Kaack,‌​‌ L.Lenneke Kuijer,​​ A.-L.Anne-Laure Ligozat,​​​‌ J. T.Jan Tobias‌ Muehlberg, Y.Yunmook‌​‌ Nah, T.Thomas​​ Olsson, A.-C.Anne-Cécile​​​‌ Orgerie, D.Daniel‌ Pargman, B.Birgit‌​‌ Penzenstadler, T.Tom​​​‌ Romanoff, E.Emma​ Strubell, C.Colin​‌ Venters and J.Junhua​​ Zhao. Climate Change:​​​‌ What is Computing’s Responsibility?​2025, 1-18HAL​‌DOIback to text​​
• 41 inproceedingsS.Simon​​​‌ Lambert, V.Vladimir​ Ostapenco, L.Laurent​‌ Lefèvre, E.Eddy​​ Caron, A.-C.Anne-Cécile​​​‌ Orgerie, B.Benjamin​ Fichel and R.Rémi​‌ Grivel. Revisiting virtual​​ machine consolidation to save​​​‌ resources and energy in​ heterogeneous production cloud infrastructures​‌.ICT4S 2025 -​​ International Conference on Information​​​‌ and Communications Technology for​ SustainabilityDublin, IrelandJune​‌ 2025, 1-1HAL​​back to text
• 42​​​‌ miscA.-C.Anne-Cécile Orgerie​. Code de conduite​‌ - Journées scientifiques.​​Rennes, France2025,​​​‌ 1-3HALback to​ text
• 43 miscG.​‌Guillaume Pierre. Fog​​ computing, from smart city​​​‌ services to natural environment​ monitoring.Natal, Brazil​‌May 2025HALback​​ to textback to​​​‌ text
• 44 inproceedingsO.​Ophélie Renaud, N.​‌Nicolas Gac, F.​​François Orieux and C.​​​‌Cedric Dumez-Viou. SimSDP,​ a Rapid Prototyping tool​‌ for Radio Astronomy: From​​ NenuFAR Experiments to SKAO-Scale​​​‌ Simulation.SaLF 2025​ - 11th Annual conference​‌ on Science At Low​​ FrequenciesOrleans (France), France​​​‌December 2025, 1-1​HALback to text​‌

Scientific popularization

• 45 inbook​​A.-C.Anne-Cécile Orgerie.​​​‌ Consommation électrique et efficacité​ énergétique des infrastructures de​‌ calcul.Le calcul​​ à découvertCNRS éditions​​​‌January 2025, 248-249​HALback to text​‌

Software

• 46 softwareH.​​Henri Casanova, A.​​​‌Augustin Degomme, A.​Arnaud Giersch, A.​‌Arnaud Legrand, M.​​Martin Quinson and F.​​​‌Frédéric Suter. SimGrid​.4.0April 2025​‌ lic: GNU Lesser General​​ Public License v2.1 or​​​‌ later.HALSoftware​ HeritageVCSback to​‌ text