Experience
Positions & education
-
2026 - present
Post-doctoral fellow
Inria, Bordeaux, France
Team TADAAM · with Francieli Boito
Leveraging I/O traces from production systems for I/O analysis: classification, identification of temporal patterns and outliers.
-
2024 - 2025
Post-doctoral fellow
University of Chicago & Argonne National Laboratory, Chicago, USA
Globus Labs · with Ian Foster and Kyle Chard
Co-supervision of PhD and undergraduate students, mixed-criticality scheduling, energy and carbon incentives for HPC users, resilient distributed storage with erasure coding, LLM serving on HPC systems, fine-grained parallelism on many-core systems.
-
2020 - 2023
PhD in Computer Science
ENS de Lyon & Inria, Lyon and Bordeaux, France
Teams ROMA & STORM · with Loris Marchal and Samuel Thibault
Scheduling Under Memory Constraints in Task-Based Runtime Systems: schedulers that order data-sharing tasks across multiple GPUs to minimize transfers over the memory bus, with a custom eviction policy, implemented in the StarPU runtime.
-
2022 · 3 months
Research stay
Uppsala University, Ångström Laboratory, Uppsala, Sweden
Division of Scientific Computing · with Elisabeth Larsson and Carl Nettelblad
Input-file-aware job scheduling: distributing load across compute nodes while reusing input files already loaded in a node's memory.
-
2020 · 4 months
Research internship
Inria, Bordeaux, France
Team STORM
An offline scheduler reducing data transfers in the StarPU runtime.
-
2018 - 2020
Master's degree in Algorithmics and Modeling at the Interface of Sciences
Université de Versailles Saint-Quentin, Versailles, France
Graph algorithms, game theory, distributed algorithms, databases, ranking methods, simulation and networks.
-
2015 - 2018
Bachelor's degree in Computer Science
Université de Versailles Saint-Quentin, Versailles, France
Summer schools: CSCS Summer School, Lugano (2023) · HiPEAC ACACES International Summer School, Fiuggi (2021 and 2022).
Teaching
| Course | Role | Audience | Year |
|---|---|---|---|
| Applied Algorithms | Co-instructor & project designer | Master (M2), Univ. of Bordeaux | 2026 ongoing |
| Network Programming (C project) | Teaching assistant, co-instructor & grader | 2nd-year engineering, Enseirb–Matmeca | 2021, 2022 |
| Algorithms | Co-instructor | 1st-year engineering, Enseirb–Matmeca | 2022 |
| Computer Architecture | Teaching assistant & co-instructor | 1st-year engineering, Enseirb–Matmeca | 2022 |
| Operating Systems | Teaching assistant | 1st-year bachelor, Univ. of Lyon 1 | 2021 |
| Internship supervision & defense | Internship tutor & committee member | 2nd-year engineering, Enseirb–Matmeca | 2022, 2023 |
Tutorial: StarPU / DARTS tutorial at the CONCACE workshop, Inria Bordeaux (2026). Materials.
Co-supervision & Co-mentoring
PhD students
- Alok Kamatar University of Chicago
Energy- and carbon-aware credit allocation for HPC: user incentives, simulation and a user study (SC'25). - Haochen Pan University of Chicago
Event-driven architectures for distributed scientific computing; Icicle metadata indexing and MCP servers for HPC. - Wenyi Wang University of Chicago
Fine-grained parallelism through dynamic load balancing on multi-socket many-core systems (IPDPS'26). - J. Gregory Pauloski University of Chicago
TaPS, a task-based library for evaluating execution environments (best paper, eScience'24). - Dante D. Sanchez-Gallegos Universidad Carlos III de Madrid
Wide-area data distribution over heterogeneous storage with erasure coding (CCGrid'25). - Shu Shi University of Chicago
Multi-LLM serving for HPC: placement and reuse of models already loaded in GPU memory. - Haibin Lai Southern University of Science and Technology
Task-based parallelism for LLM inference on multi-socket many-core systems.
Undergraduate students
- Karaen Senthilkumar University of Chicago
Fairness-aware hierarchical scheduling for bursty HPC workloads. - Sicheng Zhou University of Chicago
Workload resilience across task hierarchies in task-based frameworks (CCGrid'25).
Engineers
- Abdraman Mahamat
Cross-system prediction of file stripe count to improve I/O performance and data placement.
Service
Program committees
- SuperComputing 2026: Algorithms track
- SuperComputing 2025: reproducibility committee
- SuperComputing 2024: reproducibility committee
- ICDCS 2025
- EuroPar 2025
- HiPC 2025
- ISPDC 2025
- APDCM 2025
- HotCloudPerf 2025
Chairing & volunteering
- ISC High Performance 2026: session chair, Energy-Efficient Systems
- COMPAS 2026: session chair, I/O
- SuperComputing 2024: SIGHPC information booth
- IPDPS 2021: volunteer team
Reviewing
- Future Generation Computer Systems: 2024, 2025
- IEEE Cluster: 2024, 2025
- REPRO-HPC workshop: 2025
Awards & grants
-
2026
Inria–LNCC Initiative on AI €5,000
France–Brazil mobility grant for a research stay at the National Laboratory for Scientific Computing (LNCC), Brazil, with Prof. Carla Osthoff. -
2024
Best Paper Award, 20th IEEE International Conference on e-Science
One of 30 accepted papers out of 80 submissions, and the only one of three nominees to receive the distinction. -
2024
France and Chicago Collaborating in the Sciences (FACCTS) $14,000
Grant I wrote for a project with Ian Foster, Loris Marchal, Kyle Chard and Frédéric Vivien. -
2024
SIGHPC Early Career Professionals Travel Grant $600 + registration
One of only six grants awarded that year.