PhD in Dynamic Embedded AI for On-Device Neural Networks

Karlstad University; Leuven; Sur place; EUR 22 000 - 29 000; Plein temps

Avantages offerts par ce poste

PhD scholarship

Competitive salary

Flexible working conditions

Conference opportunities

Résumé du poste

KU Leuven seeks a highly motivated PhD researcher to join the Embedded Systems unit of EAVISE in Leuven, working on dynamic neural networks for embedded AI. The project explores on-device compression and adaptation of neural networks after deployment, aiming at energy-efficient computing and hardware-software co-design.

You will contribute to runtime-reconfigurable kernels, performance models, and memory-aware deployment on RISCV multicore and heterogeneous platforms.

Qualifications

• Candidates hold a master’s degree in electrical or computer engineering (or equivalent).
• Strong programming experience (C/C++) in embedded field and software engineering/compilers is desired.
• Background in computer architectures and embedded platforms (ARM Cortex-M, NPU, FPGA, embedded GPU).
• Experience with deep learning/AI is a plus.
• Excellent English and strong communication skills are required.

Responsabilités

• Engage in hardware-software co-design for embedded AI systems.
• Develop runtime-reconfigurable neural network software kernels.
• Contribute to on-device optimization and energy-efficient computation.

Connaissances

C/C++ programming

Embedded systems

Computer architectures

Deep Learning

English proficiency

Team collaboration

Formation

Master’s degree in electrical or computer engineering (or equivalent)

Outils

ARM Cortex-M

FPGA

Embedded GPU

NPU

Description du poste

KU Leuven seeks a highly motivated PhD researcher to join the Embedded Systems unit of EAVISE in Leuven, working on dynamic neural networks for embedded AI. The project explores on-device compression and adaptation of neural networks after deployment, aiming at energy-efficient computing and hardware-software co-design.

You will contribute to runtime-reconfigurable kernels, performance models, and memory-aware deployment on RISCV multicore and heterogeneous platforms.