Energy-Efficient AI Accelerator Architect (Approx. Computing)

Inria Rennes; Rennes; On-site; EUR 20,000 - 31,000; Full time

Job summary

Inria Rennes invites applications for a PhD researcher position focused on energy-efficient AI accelerators using approximate computing and hardware-software co-design.

You will investigate adaptive approximation techniques, develop modeling and prototyping workflows, and collaborate with a team to balance performance and accuracy. Strong HW/SW background and a Master’s degree are required; fluent English; Rennes, France.

Qualifications

• Master’s degree or equivalent in computer engineering or related field.
• Strong HW design and FPGA/ASIC flow experience.
• Proficient in C/C++, Python and Linux scripting.
• Familiarity with deep learning frameworks (PyTorch/TensorFlow).
• Experience with approximate computing techniques is a plus.

Responsibilities

• Investigate energy-efficient AI accelerators using approximate computing.
• Develop hardware-software co-design approaches.
• Model, simulate, and prototype with AI workloads.
• Collaborate with the team to balance performance and accuracy.

Skills

VHDL/Verilog

HW synthesis

Computer architecture

C/C++

Python

Linux scripting

ML frameworks

Approximate computing

Analytical skills

Communication

English proficiency

Teamwork

Open mindedness

Master’s degree requirement

Education

Master’s degree in Computer Engineering or related areas relevant to the PhD topic

Tools

PyTorch/TensorFlow

Job description

Inria Rennes invites applications for a PhD researcher position focused on energy-efficient AI accelerators using approximate computing and hardware-software co-design.

You will investigate adaptive approximation techniques, develop modeling and prototyping workflows, and collaborate with a team to balance performance and accuracy. Strong HW/SW background and a Master’s degree are required; fluent English; Rennes, France.