Scholarly profile
Publications & research writing.
Scientific preprints, technical reports, and thesis research on high-performance computing, numerical CFD modernizations, execution tracing, and parallel systems.
Peer-accessible technical reports, reproducible computational benchmarks, and presentation slide decks. All primary research artifacts adhere to open access and FAIR scientific data standards.
01 /Technical Reports & Archival Preprints
REPRODUCIBLE RESEARCH ARTIFACTS WITH SOURCE CODE & INSTRUMENTATION TRACES
- 012026TECHNICAL REPORT / PREPRINT
Reconstructing Execution States and Message Causality for an MPI Token Ring in Eclipse Trace Compass
Author's personal copy - Technical Report, September 2026· DOI: 10.13140/RG.2.2.22552.97284
Presents a formal methodology and instrumentation workflow for reconstructing distributed execution states and message causality in MPI-based ring topologies. Implements custom state providers, state history trees, and causality event extraction within Eclipse Trace Compass, enabling microsecond-precision analysis of token transmission delays, synchronization stalls, and inter-process communication overhead.
MPI TracingEclipse Trace CompassDistributed SystemsMessage CausalityHPC ObservabilityCITE RECORD (BIBTEX)
@techreport{ezzouine2026mpi, author = {Ez-Zouine, Ahmed}, title = {Reconstructing Execution States and Message Causality for an MPI Token Ring in Eclipse Trace Compass}, institution = {Faculty of Sciences, Mohammed V University in Rabat}, year = {2026}, month = {September}, doi = {10.13140/RG.2.2.22552.97284}, url = {https://doi.org/10.13140/RG.2.2.22552.97284} }
02 /Academic Theses & Forthcoming Monographs
Explore Architecture DossierCOMPREHENSIVE NUMERICAL CFD ENGINE MODERNIZATION (UM5R & UM6P)
Architectural Modernization and Parallel Portability of High-Order Reactive Navier-Stokes Solvers on HPC Supercomputers
Investigates the systematic modernization of IZEM, a high-order finite-difference solver for compressible reactive multi-species flows. Re-architects monolithic legacy Fortran stencils into a 4-layer modern C++17 hierarchy integrating Cantera thermodynamic kinetics, WENO-5/TENO shock-capturing schemes, parallel HDF5 I/O, and multi-node MPI execution evaluated on Morocco's national Toubkal HPC cluster.
03 /Presentations, Colloquia & Technical Talks
TECHNICAL SEMINARS, HPC WORKSHOP SLIDES & LECTURE MATERIALS
Core Concepts in Parallel Systems, Modern C++ & Kokkos
Context: Technical Seminar & HPC Systems Tutorial
A structured technical seminar and slide deck analyzing parallel computing paradigms, multi-threaded execution with OpenMP, distributed message-passing with MPI, and performance-portable abstractions with Kokkos. Examines hardware cache hierarchies, roofline performance modeling, and SLURM workload orchestration on HPC supercomputers.
04 /Open Science & Reproducibility Commitment
FAIR DATA PRINCIPLES, OPEN SOURCE KERNELS & CTEST REGRESSION HARNESS
Reproducible Computational Science
In accordance with FAIR (Findable, Accessible, Interoperable, Reusable) scientific research principles, computational kernels, tracing tools, and regression validation harnesses developed across my research projects are version-controlled and released under permissive open-source licenses. Every benchmark and numerical result reported is backed by automated CMake/CTest test suites for independent verification.