Summary
Dounia Khaldi is a Senior HPC Deep Learning Compiler Developer with 12+ years building compiler optimizations and parallel programming models for next-generation heterogeneous systems. She combines academic rigor—as a PhD from MINES ParisTech and former research assistant professor at Stony Brook—with industry impact at Intel, where she leads work on OpenSHMEM/OpenMP, LLVM parallel IR, and tensor/SPIR-V paths for DPC++. Dounia specializes in reducing data motion across complex memory hierarchies (HBM, SPM, DDR, GDDR) to deliver performance portability across CPUs and GPUs, and she has steered NSF/DOE and NVIDIA-funded projects toward practical compiler and runtime innovations. Known for translating research into production compiler tooling, she is as comfortable designing programming models as she is implementing high-level optimizations for exascale and deep learning workloads.
11 years of coding experience
10 years of employment as a software developer
Master of Science (MS), Computer Science, majoring in HPC, Master of Science (MS), Computer Science, majoring in HPC at Université de Versailles Saint-Quentin-en-Yvelines
Engineering Degree, Computer Science, Engineering Degree, Computer Science at Ecole nationale Supérieure d'Informatique (ESI, ex INI)
Ph.D., Computer Science, Ph.D., Computer Science at MINES ParisTech
French, English, Arabic