/GPU Architecture Researcher

GPU Architecture Researcher

Research & development - Leuven | More than two weeks ago

Envision new architectural and micro-architectural ideas for general purpose GPUs to enable efficient large scale future computer systems.

Apply

GPU Architecture Researcher

What you will do

The Compute System Architecture (CSA) unit at imec desires to build zeta-scale AI/HPC hardware and software solutions co-designed. We are backed by a broad in-house R&D expertise, creating a new AI computing paradigm that will move the industry forward for many years to come. We are looking for talented researchers and engineers with varied skillsets advancing the art of high-performance system codesign. Ideal candidates will channel their creativity to contribute in cross-spectrum flow and methodology, working closely with a team of software researchers, micro-architects and circuit designers.

We are continuously innovating across the stack from applications to architecture to technology, to build some of the most complex and best performing compute systems in the world. In this role, you apply your skills to develop and maintain functional and performance models of GPGPUs. You architect, design, develop and document tools to analyse and simulate key workloads of interest to validate and study new features. You grow into an authority on critical aspects of the GPGPUs and identify key bottlenecks, helping to understand scaling limitations, and comparing state-of-the art approaches in GPGPUs, for machine learning training workloads. This is your opportunity to build the next generation AI hardware at scale.

What we do for you

We offer you the opportunity to join one of the world’s premier research centers in nanotechnology at its headquarters in Leuven, Belgium. With your talent, passion and expertise, you’ll become part of a team that makes the impossible possible. Together, we shape the technology that will define the society of tomorrow.

We are committed to being an inclusive employer (http://www.imec-int.com/en/careers#diversity) and proud of our open, multicultural, and informal working environment with ample possibilities to take initiative and show responsibility. In everything we do, your future colleagues are guided by the imec values of passion, excellence, connectedness and integrity. We commit to supporting and guiding you in this process; not only with words but also with tangible actions. Through imec.academy, 'our corporate university', we actively invest in your development to further your technical and personal growth.

We are aware that your valuable contribution makes imec a top player in its field. Your energy and commitment are therefore appreciated by means of a market appropriate salary with many fringe benefits.

Who you are

Essential skills

Master’s with minimum 2 years of experience or Doctoral degree in Computer Engineering or related field.
Experience with (micro)benchmarking and performance analysis on GPUs or emerging ML accelerator architectures.
Familiar with computer architecture, both at system level and microarchitectural level.
Strong programming ability in C, C++. Scripting in Perl and/or Python.
You have an analytical mindset and think at an abstract level, yet you have a hands-on attitude and like to dive deep into a problem.
Constructive attitude, able to convince others and always prepared to listen and learn from others when solving problems.

A great candidate is familiar with one or more of these:

Good understanding of modern ISA and microprocessor implementation techniques.
Deep understanding of systems architecture: CPU, GPU, memory subsystem, kernel internals would be helpful.
Experience in performance analysis of sophisticated systems.
Experience with chip and/or system simulation.
Successful product development experience in performance modelling and analysis is a big plus.
Experience with design of accelerator interfaces and innovative architectural features for compute cores.
Experience with the use, creation or managing of simulation products or projects.

Apply