PhD opportunities

Methodology for Efficient Implementation of Artificial Intelligence Kernels

Thesis proposal

Area of expertiseReal-time computer science, robotics, systems and control - Fontainebleau
Doctoral SchoolSystems Engineering, Materials, Mechanics, Energy
SupervisorM. Claude TADONKI
Research unitMathematics and Systems
Contact
Starting dateOctober 1st 2021
KeywordsAI, HPC, parallelism, multicore, GPU, acceleration
AbstractCurrent and future expectations from AI are very challenging, and significant advances has been made in the topic. Powerful AI methods (e.g., Deep Learning) and large-scale datasets (e.g., massive networks) processing are computationally expensive, hence the need for high performance computing.
The aim of this PhD is to investigate systematic approaches to reach HPC implementation of important AI kernels considering both standard and specific devices. Precisely, shared memory parallelism on multicore processors and accelerated computing with GPUs will be the main considerations.
ProfileMaster 2 (research) and
- Good level in programming (C)
- Background in algorithm design and analysis
- Basic skills in parallel computing
FundingConcours pour un contrat doctoral